MUCH of the literature on exchange rate theory concentrates on the extreme cases of fixed or flexible exchange rates. When a government fixes the price of its currency to the currency of another country, it does so by offering to purchase or sell its own currency for the other currency at the established parity value. It has long been known that, under these conditions, the financial fluctuations in the two countries will be linked and so, as a consequence, will many other macroeconomic variables. Flexible exchange rates were once thought of as a mechanism for limiting this interdependence. If the central bank was not obliged to purchase or sell foreign currency on demand, then it would be able to independently set its own monetary and financial policies. The conclusion, then, is that macroeconomic interdependence would be reduced in a flexible exchange rate regime.


MUCH of the literature on exchange rate theory concentrates on the extreme cases of fixed or flexible exchange rates. When a government fixes the price of its currency to the currency of another country, it does so by offering to purchase or sell its own currency for the other currency at the established parity value. It has long been known that, under these conditions, the financial fluctuations in the two countries will be linked and so, as a consequence, will many other macroeconomic variables. Flexible exchange rates were once thought of as a mechanism for limiting this interdependence. If the central bank was not obliged to purchase or sell foreign currency on demand, then it would be able to independently set its own monetary and financial policies. The conclusion, then, is that macroeconomic interdependence would be reduced in a flexible exchange rate regime.

MUCH of the literature on exchange rate theory concentrates on the extreme cases of fixed or flexible exchange rates. When a government fixes the price of its currency to the currency of another country, it does so by offering to purchase or sell its own currency for the other currency at the established parity value. It has long been known that, under these conditions, the financial fluctuations in the two countries will be linked and so, as a consequence, will many other macroeconomic variables. Flexible exchange rates were once thought of as a mechanism for limiting this interdependence. If the central bank was not obliged to purchase or sell foreign currency on demand, then it would be able to independently set its own monetary and financial policies. The conclusion, then, is that macroeconomic interdependence would be reduced in a flexible exchange rate regime.

To a certain extent, the lessons from the theoretical analysis have been supported by the experience of the last decade of flexible exchange rates. It has, for example, been possible for interest rates on debt denominated in the major reserve currencies to diverge by over 10 percent and differences in national inflation rates have been far greater than they were under the Bretton Woods system. On the other hand, much of the evidence from the floating rate period conflicts with the predictions of simple exchange rate theories: exchange rates have not simply adjusted to offset inflation rate differentials; interest rate differentials have not reflected exchange rate expectations; and, most importantly, the countries participating in the flexible exchange rate system have experienced strong common movements in their economic performance. After a decade of working within the new international monetary system, most governments retain a flexible exchange rate because they have to rather than because they want to.

The purpose of the conference on exchange rate regimes and policy interdependence was to explore the channels by which economic disturbances in one country are transmitted to other countries in the post-Bretton Woods system. In the first paper, Mr. Dornbusch synthesized the theoretical contributions that had been made to our understanding of the mechanisms by which international financial markets were brought into equilibrium. He then discussed the channels through which disturbances were transmitted and reviewed the policies that might be adopted to counteract the less desirable features of the transmission process. In the second paper, Mr. Branson considered the structure of the labor, commodity, and financial markets of the industrial world, and he described how changes in the structure of these markets influenced the degree of interdependence when the exchange rate was flexible. In the third paper, Mr. Swoboda examined the evidence on interdependence between the United States and the member countries of the European Monetary System and he discussed the extent to which the formation of a currency union reduced the exposure of individual countries to externally generated disturbances. In the final paper, Mr. Arriazu discussed issues of interdependence from a Latin American perspective. He placed particular emphasis on the sectoral impact on Latin American countries of developments in international financial markets. The conference concluded with a round-table discussion relating to issues of trade liberalization in Latin America.

The purpose of this overview chapter is to provide a convenient summary of the papers presented at the conference and to impart some of the flavor of the discussions that followed the presentations. In undertaking this task, no attempt was made to critically evaluate either the papers or the discussion.

Flexible Exchange Rates and Interdependence, by Rudiger Dornbusch

During the 1960s, flexible exchange rates had been viewed as a way of combining independence of demand management policies with a liberal trade and payments system. However, experience since the breakdown of the Bretton Woods system had undermined confidence in this view: policies aimed at achieving domestic stability had produced large movements in exchange rates when practiced by small or open economies and had interfered with stability abroad when exercised by large economies. In his paper, Mr. Dornbusch reviewed the channels along which the effects of policies in one country affected economic behavior in other countries, and considered ways in which it might be possible to combine domestic policy independence with an open world economy.

Mr. Dornbusch first constructed a macroeconomic model in which assets were perfect substitutes for each other once anticipated exchange rate changes had been taken into account and in which expectations were rational. He then considered how key variables in the model might react to two separate disturbances—a rise in the foreign rate of interest and a reduction in the demand for exports. In the typical case, both caused the currency to depreciate in real terms and unemployment to rise. They also caused, as long as money wages increased, the exchange rate to overshoot its new equilibrium level. The overshooting was greater the stickier were real wages, the lower was the elasticity of foreign trade with respect to the exchange rate, and—for the first disturbance—the greater was the responsiveness of aggregate demand to a rise in the rate of interest.

Mr. Dornbusch then relaxed the assumption of perfect substitutability among assets, thereby introducing a risk premium into the model. The risk premium established a link between the distribution of world wealth, on one hand, and the interest rate and/or the exchange rate, on the other. It thus provided a channel through which the current account could affect the exchange rate. For example, a current account surplus increased home wealth, leading to a rise in the interest rate, an appreciation of the currency, or both. Mr. Dornbusch noted, however, that the current account was likely to be a relatively minor source of changes in wealth. He also pointed out that the risk premium had implications for intervention policy. By giving importance to the composition of domestic assets, it caused open market operations to affect the exchange rate through their impact on the level of debt, independently of their effect on the supply of money.

Mr. Dornbusch discussed three main ways in which expectations might cause the exchange rate to move away from its equilibrium level; he referred to these as “bubbles,” “pesos,” and “runs.” Bubbles occurred when holders of a currency realized that it was overvalued but expected the currency to appreciate further before its price fell. Speculation kept the bubble growing until new information caused it to collapse. The peso problem arose when investors doubted that an announced policy, such as a stabilization program, would be adhered to. By acting on these doubts the speculators might cause their prophecies to be fulfilled. Finally, exchange rates were determined, in part, by irrelevant information, which might give rise to runs on particular currencies as investors shifted their attention from one factor to another.

Turning to prescription, Mr. Dornbusch argued that the exchange rate regime should foster reasonably stable rates in the short run, as well as an equilibrium value of the currency and independence of domestic from foreign inflation in the long run. He then considered three alternatives to the present system: a return to fixed rates, more active intervention, and measures to reduce short-term capital movements.

The restoration of a fixed rate system would be rendered difficult, Mr. Dornbusch claimed, by the large differences in inflation rates among the key industrial countries, the apparent lack of willingness of the major industrial countries to abide by rules, and the absence of stable macroeconomic policies in the United States and the United Kingdom. Intervention aimed at maintaining the exchange rate within a band would encourage speculation; that directed toward dampening fluctuations relative to some notion of the equilibrium rate could cope with disturbances that were clearly financial but would run into difficulties when the disturbances were partly real or when their nature was uncertain. Also, intervention could not be used to deal with cyclical changes in exchange rates due to divergent national policies.

There were three main objections to the use of taxes to discourage international capital flows: that even if they could be made effective they (1) would still allow the transmission of effects through the current account and (2) would introduce distortions into the capital markets, and in any case they (3) would induce tax evasion and therefore would not be effective. Mr. Dornbusch argued that the first objection was the most serious as it implied that the advantages of interest equalization taxes were limited to avoiding the effects of exchange rate changes on real wages and prices. The second objection assumed that the short-term money market rate was equal to the social productivity of capital and was simply not valid. He agreed that there was some merit in the third objection but argued that tax evasion would not be significant for transitory taxes.

Mr. Dornbusch concluded that policy measures had effects on other economies regardless of the exchange rate regime that had been adopted. Nevertheless, the exchange rate regime was important because it determined the distribution of effects between changes in employment and changes in real exchange rates and inflation. The appropriate course of action, he suggested, might be to review the contribution that incomes policies—monetary rules and transitory wage controls—could make toward eliminating the adverse effects of interdependence. In view of the effect the present low level of activity had had on real wages, and the importance of locking in this effect, it might now be an auspicious time to introduce such policies.

The discussion following Dornbusch’s paper touched on many subjects, including the factors responsible for the recent international monetary instability, the path to a more stable performance by the international monetary system, the effects of the tight U.S. monetary policy on other countries, and the appropriate response of European countries to this policy.

While Mr. Dornbusch covered many of the factors influencing the recent behavior of the international monetary system, it was argued that attention should also be paid to the roles played by such factors as multiple interest rates and policy strategies. On the first of these, it was pointed out that just as the same time profile of nominal wages might be associated with different perceptions of real wage movements between buyers and sellers of labor services, so might the same nominal interest rates be associated with different perceptions of real interest rates between lenders and borrowers. In particular, movements in real interest rates might look quite different to lenders than to borrowers in the foreign trade sector. Such differences might have been important in explaining developments in the United Kingdom in the early years of the Thatcher Government. Regarding policy strategies, it was observed that not only macroeconomic variables such as output and prices but also policies of different countries were interdependent. Thus, in addition to Mr. Dornbusch’s theory of model interdependence, it was necessary to have a positive theory of strategic interdependence to explain developments in the international economy. A normative theory of strategic interdependence was also needed to permit the design of suitable rules of behavior for countries.

Among the policy issues considered were the distribution of emphasis between improving the exchange rate system and improving national policies, the merits of band proposals to improve the exchange rate system, and the use of incomes policies to improve national policies. With respect to the issue of strengthening the system or the policies, it was claimed that it might be as realistic to assume that national policies were given and direct efforts toward improving the exchange rate system as to assume that the system was given and try to improve the policies. However, it seemed more appropriate to consider a third possibility—whether or not a process of circular causation might exist between improvements in the system and improvements in policies; if so, this might be used to seek cumulative improvements in both. Band proposals received some support as a way of strengthening the exchange rate system. It was argued that opposition to these proposals was based partly on the misconception that they involved rather narrow bands placed around central rates and almost continuous intervention. Supporters preferred to use the concept, “target zones,” rather than “bands,” in order to suggest that there was both a relatively wide range within which rates might be allowed to move and that the authorities were not always intervening in the market. The target-zones approach involved continuous surveillance and occasional intervention. The latter, in contrast to the frequent practice of “leaning against the wind,” did not produce the perverse result of slowing down movements back to equilibrium rates. The wisdom of sustained reliance on incomes policies as a means of combating wage stickiness was also questioned. It was suggested that such policies might produce favorable results in the short run and, as the prevailing short-run problems were so severe, that their use should be given very serious consideration. However, it was also argued that over the longer term they might add to the rigidity of wages, thereby aggravating the problems faced by policymakers.

There was considerable interest in the implications of the tight U.S. monetary policy for the economic performance and policies of other countries. It was noted that, while the adverse effects of the policy on other countries have been well documented, not all the effects were unfavorable. The policy had caused the dollar to appreciate and real commodity prices to fall. The dollar appreciation had strengthened the competitive positions of the United States’ trading partners, with favorable results for those with initially weak current accounts, and the commodity price declines had benefited—among others—importers of oil and of U.S. wheat, corn, and soybeans. Also, the whole world gained from the progress made by the United States in reducing inflation. Regarding the appropriate European policy responses, it was argued that, if effective demand depended on the exchange rate, European countries should accept the downward movements of their currencies as a partial offset to the depressing effect of the U.S. policies on world demand. However, if exchange rate changes had no effect on real demand, as appeared to have been suggested by the implied stickiness of real wages in Europe, then the only effect of currency depreciation was to increase inflation. Europe should, therefore, adopt policies that would prevent such a depreciation.

Economic Structure and Policy for External Balance, by William H. Branson

For the first 25 years after World War II, the emphasis in macro-economic studies was on the use of one-commodity models to analyze the interactions between the real and financial sectors of the economy and to examine the role played by aggregate demand in determining the values of such variables as output and the general price level. The version of these models used in international economics was the monetary model of the balance of payments. However, the disturbances experienced by the industrial countries during the 1970s were mainly in the form of changes in relative prices. This development led to increased attention being paid to the roles of structure and supply and to the use of multi-commodity models to facilitate analysis of those roles. In his paper, Mr. Branson discussed some of the more important results of these studies.

Mr. Branson provided a framework and a starting point for his exposition by constructing a simple monetary model in which goods were perfect substitutes, wages and prices were perfectly flexible, and domestic and foreign interest earning assets were perfect substitutes. These assumptions ensured, respectively, that domestic and foreign prices were related by the exchange rate, that output was at full employment level, and that with a fixed exchange rate domestic and foreign interest rates were equal. After illustrating the workings of the model, Mr. Branson considered three policy options that the model implied for a country suffering from internal and external imbalance as a result of a budget deficit. These options were (1) the reversal of the budget deficit until the money supply had returned to its initial equilibrium level; (2) the elimination of the deficit, together with a devaluation of the currency that validated the increase that had already taken place in the money supply; and (3) an initial devaluation, followed by further, diminishing devaluations (a decelerating crawling peg) to accompany gradual elimination of the budget deficit. While the simple monetary model suggested that all three were reasonable solutions to the country’s stabilization problem, their feasibility in practice was affected to an important extent by structural factors not incorporated in that model. Mr. Branson then proceeded to modify the model and used the modified versions to examine these structural factors.

To incorporate relative price changes, Mr. Branson relaxed the assumption of a single commodity. He examined the effects of a devaluation on the trade balance of four types of economy: a small economy, in which both export demand and import supply elasticities were infinite; a semismall economy, that had some market power on the export side so that the elasticity of demand for its exports was less than infinite; a rigid economy, in which both export supply and import demand elasticities were zero, such as an economy using intermediate imports to produce a supply-inelastic export; and a pure manufacturing economy, which imported intermediate goods and exported manufactures and where the elasticity of demand for imports might be close to zero. Mr. Branson noted that it was only during the last decade that much attention had been paid to intermediate imports, which were important in the rigid and pure manufacturing economies.

In the small economy case, the devaluation had no effect on relative prices but did improve the trade balance. It worsened the terms of trade in the semismall economy case, and might improve or worsen the trade balance depending on the values of the demand elasticities; if the latter were low, the devaluation could produce J-curve effects and exchange rate instability. For the rigid economy, the devaluation had no effect on the terms of trade or on trade volumes; however, it did increase the absolute value of the trade balance in proportion to the exchange rate change, and if the initial balance was a deficit, the devaluation accordingly increased that deficit. It could also lead to a rise in unemployment because of the effect that higher prices had on real balances and that the increased trade deficit had on domestic incomes. In the pure manufacturing case, devaluation could worsen both the terms of trade and the balance of trade. The range of these results, and the possibility that devaluation might worsen the trade balance and lead to exchange rate instability, illustrated the need, Mr. Branson argued, to pay attention to trade structures when evaluating exchange rate policies.

Next, Mr. Branson relaxed the assumption of perfect wage and price flexibility in order to examine the role of the exchange rate and, to a lesser extent, that of aggregate demand when wage rigidities of various kinds were present and markup pricing was practiced. Reverting to the one-commodity model, he assumed that nominal wages were sticky in one country, say, the United States, and real wages were sticky in another area, say, Europe. In this situation, an expansion of demand in Europe, not surprisingly, increased prices in both Europe and the United States and produced a current account deficit in Europe and a current account surplus in the United States. However, because of the different nature of the wage rigidities, the demand expansion increased output only in the United States; in Europe, the only domestic effect was the rise in prices. In view of this, Mr. Branson observed, it was not surprising that when the United States had supported the locomotive approach in 1977, Europe had rejected it.

In the one-commodity model, an exchange rate change also affected output only in the United States. However, this was not the case if Europe and the United States produced different commodities. In this situation, the relevant real wage for European employers was the nominal wage deflated by the prices of domestic goods, while for European workers it was the nominal wage deflated by a weighted index of the prices of imported, as well as domestic, goods. An appreciation of the dollar, which caused the nominal wage in Europe to rise to maintain the real wage that was of interest to European workers, now increased the real wage that was of concern to European employers, who responded to this development by reducing output. Mr. Branson completed this section by demonstrating that, if wages were fully indexed, an exchange rate change would produce a proportionate change in all other prices and thus would affect output only through its impact on real balances.

Turning to the role played by the structure of financial markets, Mr. Branson first converted the monetary model into a model of exchange rate determination and used this model to show how changes in the money supply could be used to bring about the exchange rate adjustments needed to stabilize domestic prices when the foreign price level was fluctuating. He then relaxed the assumption that domestic and foreign assets were perfect substitutes and showed that price stabilization policy must now pay attention not only to the change in the money supply but also to the manner in which the change was brought about—whether it was through open market operations or intervention. If it was through intervention, the change in the money supply needed to effect the desired change in the exchange rate was smaller than if open market operations were used.

Finally, Mr. Branson observed, the exchange rate of a floating currency was determined in the short run by conditions in the financial markets. Only if these markets were deep and well-integrated with financial markets abroad could the rate be reasonably stable. This helped explain the division of countries into those that could float successfully—mainly the industrial countries—and those for which pegging was the only feasible option.

The discussion following Mr. Branson’s paper also touched on a wide range of subjects. Among those that received the most attention were the importance of trade structure in determining the effects of exchange rate changes, the role of financial structure in influencing the feasibility and desirability of floating, the conditions required for successful pegging, the importance of recent fluctuations in current account balances and exchange rates, and some implications of the poor predictive record of exchange rate models.

The importance of trade structure, and the different elasticities it implied for various categories of exports and imports, in determining the effects of exchange rate changes on the current account balance was generally recognized. One participant illustrated this by describing the contrasting effects on the French economy of devaluation of the franc against the dollar, on the one hand, and against the deutsche mark, on the other. France’s trade in dollar-denominated commodities was weighted toward intermediate rather than final products. Devaluation of the franc against the dollar produced inflation, a worsening of the current account balance, and a decline in economic activity in France. That country’s trade with the Federal Republic of Germany, however, was mainly in final products and the two countries competed strongly for each other’s markets. Devaluation of the franc against the deutsche mark, therefore, had little immediate effect on inflation, improved the current account balance, and gave a boost to economic activity in France.

Issue was taken with the argument that devaluation reduced the balance of trade in a country that exported manufactured goods and imported intermediate goods. Japan and Korea, it was pointed out, had this kind of trade structure, but both countries had used traditional exchange rate changes to correct balance of payments disequilibria—revaluations of the yen had reduced Japan’s trade balance and devaluations of the won had increased Korea’s trade balance. It was suggested, however, that the cases of Japan and Korea indicated not that the conclusions about trade structures and elasticities were wrong but that other factors might have been operating in those countries that had given rise to the particular results of the exchange rate changes.

It was generally agreed that sophisticated financial markets were a necessary condition for successful floating. It was not, however, a sufficient condition. It was pointed out that, beyond a certain point, increased integration of financial markets would render them more susceptible to monetary shocks from abroad. Also, bandwagon and overshooting effects could occur if thick financial markets were accompanied by thin labor and product markets. Participants emphasized that even though financial and other conditions might make floating a feasible policy option, it still might not be the most desirable kind of exchange rate regime to adopt.

Attention was also paid to the conditions that made pegging successful, and indeed preferable, in a small country that was able to exercise some control over capital movements. First, it was stated, the monetary authorities must be able to intervene in amounts sufficient to prevent excessive fluctuations in exchange rates; to do this, they had to maintain a high average level of international reserves in order to permit the necessary purchases of the domestic currency when the latter was tending to depreciate excessively. Second, the authorities must be able to follow a flexible monetary policy and be willing, when necessary, to introduce and persevere with tight monetary policies to make absorption consistent with available supplies. Third, the government must have a strong incomes policy in order to ensure flexibility of real wages. It was easier to pursue the required intervention, monetary, and incomes policies if the country’s currency had not been widely used in international transactions.

Participants also questioned whether fluctuations in current account balances and in nominal and effective exchange rates had been excessive since the advent of floating. With respect to current account balances it was argued that such movements had not been unduly large for the industrial countries as a whole and had been smaller for the larger countries than for the smaller countries that pegged their currencies. It was also noted that fluctuations had not been as large since 1974 as they had been in the 1920s and 1930s. Regarding exchange rates it was suggested that, while fluctuations in nominal rates had been larger than had generally been anticipated before the introduction of floating, movements in real rates had been reasonably moderate. Further, fluctuations in exchange rates were not, per se, undesirable. The purpose of having a flexible rate regime was to allow rates to vary. This permitted, among other things, countries to pursue independent fiscal and monetary policies. The conditions that would produce stable rates under a flexible rate regime, including the willingness of countries to assign to monetary policy the role of protecting the exchange rate, would also render unnecessary that kind of regime—they would permit the adoption of a fixed rate system. In this context it was suggested that the wrong conclusions might have been drawn from the fluctuations that occurred during the Great Depression. Rather than aggravating the decline in economic activity, these fluctuations probably reduced its magnitude by cushioning some of the effects of the disturbances that were occurring at that time. It was clear, nevertheless, that regardless of how one assessed recent fluctuations, movements in exchange rates could be excessive. What was needed was a yardstick that would permit judgments to be made about the appropriateness of swings in rates.

Some doubt was cast on the accuracy and usefulness of the assumptions used in exchange rate models and, thus, also of the conclusions derived from these models. It was argued that, based on their predictive records, most empirical models of exchange rate behavior had been failures—in fact, they had been less successful than the spot rate itself in forecasting the exchange rate. In view of this, care should be exercised in accepting the conclusions drawn from exchange rate models, such as the proposition that increases in the money supply led to depreciation of the domestic currency. It was also argued that one possible implication of the inability of economists and exchange market analysts to predict exchange rate changes, namely, that these changes followed a random walk, should not be used to justify inclusion in exchange rate models of the assumption that nominal interest rates were the same across countries. This assumption was not only demonstrably false; using it diverted attention from the real issue, which was to explain why nominal interest rate differentials had been so large. If this question could be answered, it might be possible to narrow interest rate differentials and, through this, contribute substantially to eliminating exchange rate variability. In defense of existing exchange rate models, it was suggested that the poor predictive performance during the 1970s lay not in the assumptions but in the nature of the shocks to which economies were subjected during the period. It was also argued that the poor forecasting record was not surprising to supporters of the asset markets approach to exchange rate behavior, who inferred from it that the role played by new information dominated that of trend factors in determining exchange rate movements.

Exchange Rate Regimes and U.S.-European Policy Interdependence, by Alexander K. Swoboda

Recent theoretical analysis and empirical evidence had cast doubts on the traditional view that economic interdependence was less under flexible than under fixed exchange rates. Mr. Swoboda discussed the relationship between interdependence and exchange rate regimes, considered the implications of his conclusions for economic policy, and examined both the insulating properties of currency unions such as the European Monetary System and the merits of a return to fixed rates between the United States and Europe.

Mr. Swoboda analyzed the behavior of five major macroeconomic variables (prices, output, interest rates, money supply, and government spending) in six leading industrial countries (Canada, France, the Federal Republic of Germany, Japan, the United Kingdom, and the United States) during both the fixed exchange rate period of the 1960s and early 1970s and the flexible exchange rate period from the mid-1970s. The results showed a marked increase in the average rate of inflation and, in line with traditional thinking on interdependence, a widening of rates of inflation from the fixed to the flexible exchange rate period. The average rate of output growth also fell in the latter period but, in contrast with what traditional theory would have suggested, the decline was fairly generalized. No clear patterns emerged for the other three variables and, Mr. Swoboda suggested, there were no a priori reasons why they should.

These results shed little light on such questions as the existence of short-term or medium-term interdependence, the role of common factors in explaining variations in country series, the influence of innovations, the importance of systematic differences in the degree of association among variables as between the two exchange rate periods, and the existence of regional groupings based on interdependence. Mr. Swoboda used correlations and principal components analyses to help answer such questions. These indicated a rise in the degree of association from the fixed to the flexible exchange rate period, for changes in inflation, output, the money supply, and interest rates. They also suggested the probable existence of both a common factor in explaining fluctuations in output and of a regional grouping of countries into North America and Europe for changes in output, interest rates, and—for the fixed rate period only—the money supply.

Traditional theory suggested that, under a fixed exchange rate regime, macroeconomic target variables were highly interdependent and autonomy in the formulation of monetary policy was extremely limited. Essentially, all countries were on a common currency standard, and economic relationships between countries were, in many ways, similar to those between regions of the same country. Thus, inflation rates and interest rates tended to equality across countries and changes in economic activity in one country induced corresponding changes in other countries. Some autonomy existed for fiscal policy but, except for the reserve currency issuing countries, monetary policy was severely constrained by the need to protect the exchange rate. Under a flexible rate regime, in contrast, differences in inflation rates could be accommodated by changes in the nominal exchange rate and domestic output could be protected from the effects of fluctuations in growth rates abroad by appropriate changes in the exchange rate and demand. For these reasons, it had been widely believed, in the 1960s and early 1970s, that a move from fixed to flexible exchange rates would reduce economic interdependence and would provide greater freedom for the operation of monetary policy.

The doubts that experience since the mid-1970s had cast on the insulating properties of a flexible exchange rate regime had encouraged the study of transmission mechanisms under flexible rates. Laursen-Metzler had shown that a boom abroad could lead to a slump at home; Mundell-Fleming had demonstrated that fiscal expansion at home could produce an expansion abroad, while monetary expansion at home could lead to contraction abroad; and Mr. Dornbusch had shown that monetary disturbances could cause the overshooting of exchange rates.

However, Mr. Swoboda argued, there were two main difficulties with these approaches: they indicated mainly negative channels of transmission, and they suggested that scope was greater for offsetting the effects of foreign disturbances under flexible than under fixed rates. They did not, therefore, explain the high degrees of association of short-term inflation rates and of output growth rates under flexible rates. While the latter might be due, in part, to transmission, they might also reflect, Mr. Swoboda suggested, the existence of a world business cycle. Such a cycle could arise because shocks were specific, not to countries but to industries and stages of production across countries. In this context, he argued, it might be more useful to divide the world into industrial sectors than into countries.

In his final section, Mr. Swoboda discussed two specific questions that were addressed to him by the organizers of the seminar: how could arrangements such as the European Monetary System affect interdependence? and whether or not fixed exchange rates should be re-established between North America and Europe? The traditional view that interdependence should be strong among currency union countries but weak between these countries and those outside the union was not supported by recent experience. Mr. Swoboda suggested, however, that the formation of a currency union might, by making the monetary policies of members more predictable, reduce the variability of real exchange rates and, through this, the severity of outside shocks. With respect to the choice between fixed and flexible rates, Mr. Swoboda noted that output interdependence and U.S. dominance had been as great since the adoption of flexible rates as they had been before. Nevertheless, the large industrial countries still preferred flexible rates, mainly because these reduced interdependence of inflation levels and appeared to avoid the need for commitment to rules of monetary behavior. However, Mr. Swoboda argued, the appearance of greater policy independence was illusory. Policy commitments were required for successful working of both types of exchange rate system; it was only the form of the commitment that differed.

The discussion focused on the causes of the relative constancy of correlations among major macroeconomic variables from the 1960s to the 1970s and of the general absence of lagged relationships in these variables in the 1970s, the implications of the empirical findings on regional groupings for the feasibility of monetary unions in Europe and North America, the implications of recent developments in Europe and the United States for the usefulness of existing transmission models, and the importance of policy autonomy and policy discipline for the choice of an exchange rate regime.

Considerable interest was shown in Mr. Swoboda’s empirical finding that the correlations among major economic variables across countries were largely unchanged from the 1960s to the 1970s. Four possible explanations were suggested for this result: (1) that the differences between the regimes of the 1960s and the 1970s were not, in fact, very great; (2) that the degree of interdependence was not greatly affected by the type of exchange rate regime in effect; (3) that the shocks to which the world was subjected in the two periods were substantially different; and (4) that the data incorporated not only the effects of the differences in exchange rate regimes and shocks but also those of the policies adopted in response to the shocks. There was support for the view that the move from a fixed to a flexible rate regime probably had not significantly influenced developments; the world recessions of 1974–75 and 1981–82, for example, would have occurred whatever the exchange rate regime. However, it was pointed out, the empirical data presented did not permit this view to be tested as they did not separate the effects of the exchange rate change from those of other factors.

The data presented by Mr. Swoboda showed that movements across countries of such variables as output and prices were not lagged, as would be suggested by business cycle transmission theory, but were contemporaneous during the 1970s. It was agreed that this could have been due to the emergence of a world business cycle. However, it was claimed, there was little positive evidence to support this explanation. In particular, the increased interdependence of interest rates in the 1970s should not be cited, as this development could be attributed to the changeover from a fixed to a flexible exchange rate regime.

The findings of contemporaneous correlations could be explained more convincingly, it was argued, by the occurrence of common shocks and common policy responses to these. It was pointed out that the industrial countries had become increasingly integrated over the last three decades as a result of such factors as increased dependency on foreign trade, especially on imports of fuel and raw materials. The greater integration had fostered not only common expectations but also similar political and social policies, as evidenced by the increased government provision of goods and services and by the high employment policies of the 1960s and early 1970s. It had also fostered common structural developments such as the increase in rigidities in the labor market and the rise in the share of wages in gross national product in 1969–75. In view of these developments, it was not surprising that the large increases in oil prices in 1973–74 and 1979–80 had had a major impact on all the industrial countries and that the latter had adopted similar policies in response to them. It was pointed out, however, that the findings of Mr. Swoboda did not indicate that there had been an increased common policy response to common shocks.

It was recognized that the data presented by Mr. Swoboda had provided a basis for dividing the observed countries into a European group and a North American group. It was noted, however, that these data did not imply that a monetary union was feasible or desirable for either group. Before such assessments could be made it was necessary to consider other factors, including attitudes toward and policies for combating inflation, that determined the extent to which monetary policies within the groups could be harmonized.

While most theoretical models predicted negative transmission of the effects of foreign disturbances, especially for output and prices, such variables had tended to move in the same direction in the industrial countries in recent years. It was argued that this did not imply that the models were wrong. In fact, when policy changes were made endogenous, the models, or at least those using the asset markets approach, fitted the facts rather well. This view was illustrated by developments following the introduction of the tight money policy in the United States in 1979. As theory suggested, this policy had had a negative effect on growth and inflation in the United States and a positive effect on inflation in Europe. Because of the largely offsetting effects of currency depreciation and reduced real money supply associated with unchanged policies in Europe, there was little effect on output in Europe. This was not, however, the end of the story. European policies did not remain unchanged but were also tightened in line with those of the United States. The induced change in European policies contributed to the reduction in growth and inflation in Europe that had unnecessarily raised doubts about the usefulness of transmission models.

One of the reasons for favoring a flexible over a fixed exchange rate regime was that the former permitted autonomy in the exercise of monetary policy. It was argued, however, that this autonomy was largely illusory. Policymakers might have freedom with respect to the use of instruments but the latter might not have the potential to reach the chosen policy targets. Specifically, traditional instruments of monetary policy appeared to be ineffective now in promoting internal balance. In the light of this, one participant observed, it might be useful to assign to monetary policy the task of attaining and maintaining external, rather than internal, balance. On the other hand, it was claimed, the argument that a fixed rate regime imposed a firm discipline on the national authorities and reduced the likelihood of irresponsible policies being adopted might also be exaggerated. Rather than adapting their policies to a given exchange rate regime, governments were likely to adapt the exchange rate regime to whatever discipline they chose to impose on themselves.

Policy Interdependence from a Latin American Perspective, by Ricardo H. Arriazu

Mr. Arriazu began his paper by noting that the last few years had been extremely difficult ones for the economies of the Latin American countries. While domestic conditions and policies had undoubtedly played some role in the economic performance of the countries in the region, the general malaise, observed in conjunction with the adoption of very different macroeconomic strategies, suggested that part of the problem arose from the transmission of externally generated economic disturbances into the region. Mr. Arriazu suggested that developments in the world economy, including the expansion of international trade in both goods and capital, the adoption of flexible exchange rate and interest rate policies in the reserve currency countries, and the inflationary environment of the past decade had all contributed to the sensitivity of Latin American economies to externally generated disturbances.

Mr. Arriazu suggested that, given the variety of disturbances experienced during the past few years, it was difficult to build a complete model that would illustrate the impact of these disturbances on a typical Latin American economy. He proposed, as an alternative, to consider the transmission of a stereotyped cycle from the world economy to a stylized Latin American economy. The cycle consisted of an initial increase, and subsequent decline, in world inflation which would lead to a related cycle in world nominal and real interest rates, and the objective of his analysis was to consider the impact of this change on both the macro-economic and sectoral performance of a model with three commodity sectors—exports, import substitutes, and nontraded—and a developed banking system. The Latin American flavor of the model arose from its concern with widespread indexation of both prices and interest rates and its detailed specification of the linkages between fiscal and monetary policy and the inflationary consequences of the cycle. Mr. Arriazu’s small open economy was characterized by private and public sectors that responded endogenously to developments in the world economy.

Mr. Arriazu’s analysis of the world inflationary cycle concentrated on the relationship between the inflation of commodity prices and real and nominal interest rates. While other macro-economic variables were of obvious importance to the center countries, Mr. Arriazu felt that the main transmission channels to the Latin American economies was through price and interest rate effects. In the early stage of the cycle, higher inflation led to an increase in nominal interest rates and a possible decline in real interest rates. However, when inflation declined, nominal interest rates tended to remain high, leading to high real interest rates. In the final stage of the cycle, real and nominal interest rates were assumed to return to their longer-run, equilibrium levels.

The cycle was transmitted to the stylized economy through price and interest rate developments. Mr. Arriazu argued that the growth of world trade had led to a more rapid transmission of inflation from the center countries to the Latin American economies. His analysis of the interest rate transmission channel recognized the fact that private and public debt raised in world capital markets were tending to pay variable interest rates, so that increases in nominal interest rates were leading to additional interest costs for the countries under consideration. In the model, this factor led to a public sector deficit in the early stages of the cycle, even when indexation was complete. However, Mr. Arriazu argued, the fiscal deficit was not likely to lead to a balance of payments deficit in the early stage of the cycle because of the price-induced increase in the demand for the domestic monetary base and the associated increase in the world demand for exportables. In the second stage of the cycle, the world inflation was transmitted, through demand and indexation effects, to the prices of nontraded goods.

When the world inflation began to decline, Mr. Arriazu suggested, public sector deficits would increase sharply. Interest rates on external borrowings would remain high, and domestic tax sources would not automatically increase to offset the high borrowing costs. Given the decline in world commodity prices, in conjunction with a continued domestic response to the previous inflation, world demand for exportables would decline and local demand for import substitutes would increase. Hence, the expanding public sector deficit would, in this instance, be associated with an overall balance of payments deficit. If the prior stock of external debt was sufficiently large, the combination of a public sector deficit with an overall deficit on the balance of payments might lead to concerns over creditworthiness.

Mr. Arriazu suggested that the scenario described in the stereotype model was broadly consistent with the response of Latin American economies to the developments in the international financial system during the past few years. Despite the great differences in the economic policies followed by Latin American nations, the importance of developments in the rest of the world and the increase in the sensitivity of Latin American countries to external developments had led to a number of common features in their recent economic development.

In the final section, Mr. Arriazu briefly considered the modifications that would be required if his model was restated with a flexible exchange rate regime. Mr. Arriazu defined a flexible exchange rate regime as one in which the monetary base was exogenous, so that the exchange rate adjusted to ensure that the demand for the base was equal to the exogenous supply. While a flexible exchange rate regime exacerbated relative price movements during the cycle, and while it consequently created greater volatility in output and employment, the basic lessons from the earlier analysis continued to hold true under flexible exchange rates. Since Latin American debt remained characterized as being denominated in dollars and paying a flexible interest rate, an interest rate cycle would continue to exert real effects on the economy through variations in interest payments on public sector debt.

The discussion of Mr. Arriazu’s paper centered on two issues. The first was the extent to which the 1980–81 experience was a good representative case for illustrating the relationship between the Latin American economies and the rest of the world. The second was the extent to which the model’s assumption of fixed exchange rates limited its generality.

As far as the representativeness of the 1980–81 experience was concerned, Mr. Arriazu had suggested that the common experience of the Latin American countries was due to the transmission of world shocks through price and interest rate developments. In the discussion, the point was made that major changes in the world economy had taken place when the economic performance of the Latin American countries was extremely divergent. As an alternative interpretation of recent experience, it was suggested that the correlation in the recent performance of Latin American countries might have been due to their similar domestic policy response to the fall in real interest rates in the middle 1970s. It was apparent that many countries borrowed more than they would have borrowed had they known about the future evolution of real exchange rates and real interest rates. If the 1980–81 experience had simply been the result of expectational errors, it might not be an appropriate case for illustrating future relationships between the Latin American group and the world economy.

In regard to the question of the exchange rate regime, a number of the participants observed that Mr. Arriazu’s analysis relied too heavily on the assumption of a fixed exchange rate regime. While the response of the economy to an external shock might be similar, in the long run, for both pure fixed and pure floating rate regimes, most Latin American countries did not have pure systems and the distortions introduced by trade and capital controls, over-valuation or undervaluation of the exchange rate, and inappropriate monetary and fiscal policies might have an important role to play in assessing the impact of an external shock on a Latin American economy.

Panel Discussion on the Southern Cone

Panel Discussion with Jacob A. Frenkel, Rudiger Dornbusch, William H. Branson, and Ricardo H. Arriazu.

Mr. Dale opened the discussion of liberalization in the Southern Cone by noting that three Latin American countries—Argentina, Chile, and Uruguay—had, since the mid 1970s, adopted stabilization programs that had included moves toward more liberal exchange rate and trade policies in conjunction with preannounced exchange rate targets. Mr. Dale also noted that these three countries had also encountered difficulties in recent times, and that the issues surrounding liberalization were of great interest, particularly at the present time.

Mr. Frenkel began the discussion by noting that it was generally better to liberalize commodity flows before capital flows, and that it was important that domestic policies be corrected before any liberalization of the foreign sector was implemented. If liberalization was attempted before budgets were balanced and money growth controlled, then it was unlikely to be anything but a temporary phenomenon. The main reason why the goods markets should be liberalized before the capital markets was that goods markets tended to clear more slowly than capital markets. Hence, an external balance achieved through the trade account would be more gradual, and hence less disruptive, than an external balance achieved through capital account liberalization.

Mr. Frenkel subsequently noted that liberalization was often part of a total stabilization package. As far as the stabilization package was concerned, he argued that a rapidly imposed stabilization package would probably be more effective than a gradually imposed one, because a government needed to have a well-defined and consistent change in policy in order for the stabilization package to be credible. If the government demonstrated its credibility through its monetary and fiscal policies, it would find it easier to negotiate incomes policies and capital controls in order to smooth the private sector adjustment.

While Mr. Frenkel recognized that arguments could be made for income policies and capital controls during a transition period, he recognized that these policies could be counterproductive if the transition period did not end because of the failure of the underlying stabilization policies.

Mr. Dornbusch continued the discussion of stabilization policies by noting that the periods of economic crisis in Argentina, Chile, and Brazil were associated with high real interest rates and overvalued currencies. Mr. Dornbusch argued that the source of the problem was the failure to associate a fiscal reform with a real depreciation of the currency. In many cases, he suggested, central banks had used financing to delay adjustment in real exchange rates and had, consequently, established the conditions for a financial crisis. Mr. Dornbusch suggested a return to the more traditional approach to stabilization; an approach based on correcting budget deficits, a substantial real depreciation, and a controlled capital account program.

In his discussion, Mr. Branson returned to the problem raised by Mr. Frenkel concerning the choice between liberalizing trade in goods before liberalizing trade in capital. He suggested that the elimination of capital controls would generate an increase in confidence that would lead to an inflow of capital into the country. The capital inflow would be associated with an appreciation of the real exchange rate and a decline in national competitiveness. However, liberalizing trade would, he suggested, lead to a surge in imports, a current account deficit, and a depreciation of the currency. There were, he concluded, problems to be faced with either approach to liberalization.

Mr. Arriazu began his discussion of the issue of whether to liberalize commodity flows before trade flows by distinguishing between opening the economy to trade and reducing the level of effective protection. In a situation in which there was “water in the tariff,” the commodity was so heavily protected that even a substantial increase in its domestic price would not lead to imports of the commodity. Mr. Arriazu argued that opening the capital account in a situation in which many commodities had been removed from the traded goods sector due to excessive levels of effective protection would lead to instability in the balance of payments.

Mr. Arriazu also questioned Mr. Dornbusch’s contention that capital controls were a necessary policy instrument. In Argentina, he suggested, it was extremely difficult to maintain capital controls for any length of time because of the ingenuity with which Argentine citizens created methods for effectively evading the controls. Mr. Arriazu also felt that the role of preannounced exchange rate changes was to establish the credibility of the stabilization policy. In Argentina, he suggested that the size of the government deficit determined the extent of the necessary change in the exchange rate and that the preannouncement of the exchange rate changes provided some of the reduced uncertainty of a fixed rate regime.