I am grateful to Armando Linde and Kent Osband for helpful comments on an earlier draft. The views expressed in this paper are those of the author and in no way represent the official views of the International Monetary Fund.

In the first Carlos F. Díaz Alejandro lecture presented before a recent meeting of the ASCE, Felipe Pazos provided a lucid and comprehensive presentation of the key economic problems that are likely to face Cuba in the event of a decision to re-orientate the economy away from central planning and toward a market-based system. ¹ In concluding his lecture, Pazos expressed the hope that his thoughts would provide a framework in which the economic problems that Cuba was likely to face in the period of transition could be further discussed and analyzed. In this paper I attempt to respond to Felipe Pazos’ challenge and take the discussion of some of the important issues that he raised one small step further. In doing so, I certainly do not aim to be comprehensive. Indeed, I will deliberately leave aside some of the fundamental questions raised by Pazos (for example in the all-important areas of privatization and property rights) and concentrate on a limited number of specific policy issues, including trade policy, the exchange rate system, monetary policy and price reform.

In examining these issues, I will stress the importance of arriving at a clear understanding of the link between long-term policy objectives and the policies adopted during the transitional period. In his lecture, Pazos suggested that “we must distinguish between the basic policies that we should apply once we have overcome the problems of the transition, and those temporary measures that we will have to implement initially.” I will argue that there are some risks in applying too rigidly this principle of separation between transitional measures and long-term economic goals. Specifically, I will argue that long-term strategic objectives will need to be clearly defined and carefully explained at an early stage, and that transitional policies will need to be framed in such a way as to avoid the risk of permanent deviations from those long-term goals.

Before turning to an examination of these issues, I would like to make two related observations. First, the policies that will be implemented during and after the transitional period will be largely the outcome of a political process of decision making–hopefully one in which the entire Cuban population will be allowed to express its views freely and fully. Therefore, where this paper adopts a normative tone, the intention is to be persuasive rather than prescriptive. Second, since the paper moves into uncharted territory, its conclusions should be regarded as tentative, and its recommendations are intended to encourage, rather than to preclude, further discussion.

I. Long-Term Objectives: The Experience of Other Developing Countries

Much of the recent discussion concerning the future of the Cuban economy has focussed on how best to achieve the transition from a highly centralized form of economic planning to a market-oriented economy on the assumption that the political conditions for such a transformation would somehow emerge. This section addresses a different, albeit related, question: what is the form of economic organization and the economic policy framework that Cuba should adopt over the long term.

Reforming the Cuban economy without a clear view of the ultimate economic goals that are being pursued would involve serious risks: without a long-term strategy that serves as a guide for action in the near term economic policies could drift aimlessly, and serious mistakes could be made which may prove very difficult to correct subsequently. The basic objectives underlying the new economic policy must therefore be clearly stated by the authorities at an early stage, and these objectives must be well understood by the population. Otherwise, irrealistic expectations of a quick improvement in living standards will outrun the possibilities faced by policymakers, who initially will face severe constraints, giving rise to demands for unsustainable policies, excessive wage increases, and possibly to labor unrest. In that situation, the competitive position of the economy will be eroded, instability will hinder tourism and discourage investment by local as well as foreign residents, and the exodus of human capital–now presumably unfettered by domestic legal restrictions–will be difficult to contain.

The search for the basic principles that should govern the country’s economic strategy cannot be disassociated from the political goals pursued by society. I will assume that these goals will involve the building of a democratic and pluralistic society based on the rule of law. I will also assume that the process of economic transformation will aim at establishing an economic system capable of achieving growth on a sustainable basis, and therefore a durable improvement in the standard of living of the population. The achievement of both objectives would seem to require an evolution of the Cuban economy towards a decentralized system where private enterprise and the free operation of markets would play a significant role. But market-based economic systems can differ significantly in terms of the extent and the nature of government intervention in the economy. They also differ considerably in terms of economic performance. It is therefore important to specify in some detail the policy framework within which market forces are most likely to bring about strong and sustained growth.

Developing countries: contrasts in performance

The search for a policy framework that will achieve Cuba’s long-term economic goals must be grounded in experience. Indeed, only the lessons of experience will provide reasonable assurance that the mistakes made by other countries in the past will not be repeated, and that the actions of the successful countries will be emulated. Cuba is a developing country, and therefore it is the experience of developing countries that is most relevant. To be sure, the recent experience of the U.S.S.R. and of the countries of Eastern Europe is invaluable, but this experience relates more closely to the tactical considerations associated with transition than to the strategic considerations relating to the definition and the pursuit of long-term economic goals. Those considerations require an examination of the growth performance of developing countries.

Over the past two decades, there have been dramatic differences in the growth performance of various groups of developing countries. As illustrated in Chart 1, the developing countries of Asia have experienced strong and sustained growth of per capita GDP during the period 1970-90, while the performance of other regions has been, to say the least, disappointing. ² It is not easy to generalize about the precise reasons for these differences, as experience has differed considerably among countries, even within a particular region. However, the sharp contrast between the successful Asian countries and the other developing countries raises a fundamental question about the relation between public policies and the observed differences in economic growth. While many aspects of this question remain unanswered, a growing body of empirical evidence suggests that there is indeed a strong link between economic policies and growth performance. Specifically, the more successful countries generally have pursued outward-oriented trade policies and maintained relatively unrestricted foreign trade systems; they have followed prudent fiscal and monetary policies and avoided very high rates of inflation; and they have generally limited government intervention in markets.

Saving and investment

Most importantly, the successful countries have attained high rates of investment and saving. The evidence in this area is quite strong. For example, the average contribution of capital formation to the growth of output in the 1980s has been more than twice as large in Asia as in the Western Hemisphere and the Middle East and four times larger than in Africa. It is also interesting to note that while the growth of multifactor productivity in the Asian countries in that period averaged roughly 23/4percent a year, it was only 1 percent in the Middle East and it was negative in Africa and the Western Hemisphere.

The importance of saving and investment for growth has been underscored by a recent study dealing with a sample of 71 developing countries. ³ The results of that study, which are summarized in Table 1 below, indicate that countries with high rates of national saving and domestic investment have had considerable higher rates of growth in per capita income than the countries with low savings and investment rates. ⁴ Also, the high savers generally have experienced lower inflation, significantly lower debt/export ratios and a considerably lower incidence of debt servicing difficulties during the debt crisis in the 1990s.

Table 1. Net Debtor Developing Countries: Saving and Economic Performance, 1983-90 

(In percent of GDP, unless otherwise noted)

	National Savings	Domestic Investment	Growth of Per Capita Income2	Inflation Rate2	Debt Export Ratio3	Proportion with Debt Servicing Difficulties in 1980's3
High savers	25.8	26.6	2.2	20.2	151	33
Moderate savers	16.2	20	-0.2	53.1	295	67
Low savers	6.9	14.1	-1.1	93.4	464	91
1Based on sample of 71 net debtor developing. Countries are divided into three groups according to the ratio of national saving to GDP. All figures are unweighted averages. 2Percent per annum. 3In percent.

The saving-investment-growth connection has strong implications for economic policies. First, macroeconomic policies must aim at achieving a general climate of stability and confidence conducive to appropriate levels of private saving and capital formation. In a situation of high inflation, volatile real interest rates and exchange rates, and large fiscal deficits, investment plans will be adversely affected because of uncertainty about the prospects for the economy and the authorities’ likely policy response. A recent study of private investment behavior in 23 developing countries found that high rates of inflation-taken as a measure of macroeconomic instability and uncertainty–had a significant negative impact on private investment. ⁵ Second, policies that discourage private saving, such as financial controls that keep interest rates below market levels, should be avoided. Third, and but perhaps most important, fiscal policy should be conducted so as to avoid the absorption of national saving by the public sector, except to finance government investments in areas such as infrastructure, provided that these investments are clearly justified by a sufficiently high social rate of return. The evidence from cross-section data suggests that in the period 1983-89 developing countries with relatively high rates of national saving rates had higher government saving rates and lower fiscal deficits than other countries.

Inflation

The empirical evidence strongly suggests that the successful countries have generally avoided very high rates of inflation. Inflation involves serious costs, many of which apply with particular force to the high inflation developing countries. High inflation is typically associated with high inflation variability, which creates uncertainty and undermines the confidence of domestic investors and foreign lenders. At very high levels of inflation, the economic horizon is shortened and financial instability disrupts economic decisions. High inflation is usually a symptom of fundamental problems in macroeconomic management: while it persists, private investors expect that the government eventually will be forced to take corrective monetary and fiscal measures. The longer these measures are delayed, the more disruptive the effect on the economy. It is therefore not surprising that countries with low inflation tend to have sharply higher investment ratios and growth rates of output and exports than the high inflation countries. ⁶ This finding, of course, has important implications for the appropriate stance of monetary policy, which are discussed in some detail in Section IV.

Market distortions

Finally, there is considerable evidence that the successful developing countries generally have avoided the distortions associated with heavy-handed government interference with markets. For example, these countries have avoided financial repression (as reflected in the fact that they have consistently maintained positive real interest rates), and they have pursued outward-looking trade policies. The importance of this last factor is illustrated in Table 2, which breaks down a sample of about 40 developing countries according to their trade orientation. Clearly, the more outward oriented countries have experienced much higher growth rates of potential output in the 1980s and, significantly, considerably higher growth of total factor productivity.

Table 2. Trade Orientation and Growth (In percent a year)

	Growth of Potential GDP	Capital	Contribution of Labor	Total Factor Productivity
Strongly outward'oriented countries
1975-82	8.4	4.6	1.1	2.7
1983-89	7.7	3.3	0.7	3.7
Moderately outward-oriented countries
1975-82	4.6	2.8	1.3	0.5
1983-89	4.1	1.7	1.2	1.2
Moderately inward-oriented countries
1975-82	4	2.6	1.5	-0.1
1983-89	2.7	1.4	1.5	-0.2
Strongly inward-oriented countries
1975-82	2.3	1.6	1.6	-0.9
1983-89	2.2	0.7	1.6	-0.1
Note: All figures are unweighted averages. The classification of countries by trade orientation is based on World Bank. World Development Report, 1987.

A policy framework for the long-term

This quick review of the economic performance of the developing countries suggests that Cuba’s longrun economic strategy should be based on the following principles:

• a competitive and dynamic economy where the allocation of resources is based on the free play of market forces, except in those cases where market failure demonstrably justifies government intervention, for example in the case of environmental protection;
• a predominant role for the private sector in the production and distribution of goods and services, except as regards the supply of public goods and services such as infrastructure and social safety nets;
• an outward-looking economy, open to the free flow of international trade and capital;
• a strong fiscal policy that helps to achieve national saving and domestic investment rates sufficiently high to sustain strong growth of output over the long term;
• a tax system that is administratively simple and transparent, that avoids distorting decisions about saving and investment by emphasizing consumption taxes (rather than direct taxes on income or foreign trade), and that preserves reasonable incentives to work and to invest;
• and, last but not least, a climate of confidence and economic stability, which requires a monetary policy aimed primarily at keeping inflation under control, with adequate support from fiscal and exchange rate policies.

These, in my view, are the basic objectives that should be kept in mind in designing policies for the transitional period.

II. Price Liberalization, Subsidies and Macroeconomic Stabilization

I fully agree with Felipe Pazos that an essential element of the transition toward a market-based economy will be the liberalization of Cuba’s price structure. He rightly stresses that price liberalization will need to occur at a very early stage: “as soon as production units are privatized, or even earlier.” Of course, price decontrol will need to take place at both the consumer and producer levels. To free retail prices while maintaining restrictions on producer prices would involve a squeeze on enterprise profits–at a time when demand may be weak for various reasons–with the risk of potentially serious adverse effects on output.

Price reform will need to go hand in hand with measures to ensure the autonomy of firms from centralized control and to provide firm managers with the right to hire and fire workers and determine the desired levels of inputs and production on the basis of considerations of profitability. At the same time it will be necessary to break up existing state monopolies, to eliminate the present practice of “soft budget constraints”, and to allow unprofitable firms to experience losses or even to go bankrupt. In this connection, the introduction at a very early stage of comprehensive antitrust and bankruptcy legislation will be of the utmost importance. A thorough examination of the legal and economic principles underlying such legislations should be viewed as a top priority item on which preparatory work should begin without delay.

The dismantlement of Cuba’s distorted price structure will need to involve the elimination of a wide range of existing subsidies. This will be required not only as an integral part of the process of bringing prices in line with market conditions, but also in order to relieve pressure on a fiscal position which is already difficult and may well become even more serious in the period ahead. As indicated by Pazos, the present system of subsidies and price controls, as is generally the case in centrally planned economies, tends to artificially reduce the price of wage goods. Therefore, price reform and the elimination of subsidies will initially result in a fall in the real incomes of the majority of the population . As noted by Pazos, this would occur at a time when real incomes–which already have been reduced as a result of the cut in Soviet subsidies to Cuba and the disruption of Cuba’s trade with the members of the now extinct CMEA–might be further reduced by the economic dislocations that could take place in the wake of a change in Cuba’s political and economic system. Therefore, an important question is how to cushion the decline in real income resulting from price reform and subsidy cuts without derailing the reform process.

1. Alternative approaches to price liberalization

The experience of the Eastern European countries and the current debate in the U.S.S.R. suggest three possible solutions to this problem. These solutions are not mutually exclusive and a variety of intermediate solutions would, in principle, be possible. First, the full abolition of subsidies could be

accompanied by a system of temporary price ceilings for specific “essential” commodities, including selected food items. Of course this would lead to excess demand for those commodities subject to price ceilings, which would necessitate the maintenance of a complementary system of rationing–a system which would last as long as the price controls are maintained. Second, the termination of all price controls might be coupled with the temporary subsidization of a limited range of “essential” consumer goods, although not necessarily at pre-transition levels. Third, both price controls and subsidies might be eliminated for all goods, and the effects on the poorest sectors of the population might be cushioned by the introduction of welfare payments and a system of unemployment compensation. Unfortunately, all three methods present significant problems. The third one, however, is likely to be preferable, as it is most compatible with the principles of the long-term economic strategy. ⁷

The price ceiling cum rationing solution would introduce a significant exception to the principle of freemarket pricing, an exception that may prove hard to reverse and therefore may involve a lasting disincentive to the supply of a number of goods, including in particular some key agricultural commodities. It would also deprive the government of the political benefits of what might be one of the most popular results of price reform: the elimination of shortages and “lines” in food stores. Moreover, as argued in the following section, the retention of price controls would introduce a serious risk of rapid inflation combined with growing shortages if macroeconomic policies failed to keep the growth of aggregate demand under strict control.

The second option–combining full price liberalization with temporary subsidies for a specified list of products–would have the advantage of avoiding shortages and rationing. But it would involve a budgetary cost and would introduce distortions which, judging by the experience of those developing countries that have relied heavily on subsidies, could be politically very difficult to remove. The third option–full price decontrol and elimination of all price subsidies coupled with “safety net” payments-would also involve budgetary costs. It would, however, be most compatible with the concept of a freemarket economy where poverty issues are tackled through a system of direct government transfers rather than through subsidies and price-distorting measures.

The specific solution that will be adopted will inevitably be influenced by political considerations. Nevertheless, budgetary constraints and efficiency considerations would need to be given the importance they deserve. Thus, if price controls and/or subsidies are to be retained for some time, the list of “essential” commodities will need to be kept to a minimum, and the government should firmly and publicly commit itself to a specific schedule to phase out the transitional measures over a brief period.

If a safety net is used to cushion the temporary effects on real incomes of price decontrol and subsidy cuts, budgetary considerations strongly suggest that unemployment compensation and other welfare payments must be kept at realistic levels. In this connection, the social and economic incidence of the fall in workers’ real incomes and the rise in unemployment that will inevitably occur during the transition should not be exaggerated. As noted by Lipton and Sachs, the implications for the well-being of workers of the decline in real income resulting from price reform would be offset, at least in part, by the welfare gains stemming from the elimination of shortages and queues. ⁸ It should also be noted that the consequences for production of the rise in joblessness would be limited, because this rise would represent in part a shift from disguised to open unemployment. Of course, those who become openly unemployed would lose their salaries, which provides the rationale for unemployment compensation.

Finally, the fiscal burden of financing social safety net systems need not fall entirely on the budget. In several Eastern European countries, unemployment benefits schemes are now partly financed by employer’s contributions (e.g. in Poland and Bulgaria) or by employer/employee contributions (Hungary).

2. Price liberalization, shortages, and stabilization

There is no dispute that full price liberalization is ultimately essential to eliminate shortages and queues and to promote efficiency in the use of resources. Still, it might be argued that price liberalization should be introduced gradually so as to avoid unleashing inflationary pressures that otherwise would need to be contained through highly restrictive financial policies. In this vein, it might be argued that there is a trade off between the objectives of controlling inflation and reducing shortages in the sense that a gradual process of price decontrol will diminish the degree of stringency required on the part of financial policies. I will argue that this is a dangerous fallacy, and that a strategy of gradual price liberalization combined with loose demand policies will lead to both rapidly rising prices and to growing shortages.

This kind of situation, which actually developed in Poland in 1989, has been described as “shortageflation”. ⁹

To illustrate this proposition, consider Figure 1, where the price of goods (p) is plotted on the vertical axis and the excess aggregate demand for goods (e) on the horizontal axis. (For simplicity it is assumed that the aggregate supply of goods is fixed and that all prices are controlled.) The distance between the vertical axis and the downward sloping curve ED1 represents the excess demand for goods if the price is regulated at any level below p1*. For example, if the regulated price is 1, the shortage of goods will be indicated by s1.

If aggregate demand remains unchanged, full deregulation will bring the price level from 1 to p1*, thus eliminating the shortage. This rise in the price level could be substantial given the severity of the price distortions currently affecting the Cuban economy. Table 3 shows that in 1987 the gaps between regulated and parallel market prices for a number of important commodities were extremely large. More recent data is unavailable, but there are indications that shortages have increased for many commodities, and that the number of commodities subject to rationing has risen. ¹⁰ But the sharp increase in prices resulting from deregulation will, if monetary and fiscal policies are appropriately restrained (i.e., if the schedule ED in Figure 1 does not shift upward), involve a one-step adjustment. The measured rate of inflation will show a rise for some time, but there would be no lasting effect on the rate of inflation.

	Controlled Market	Parallel Market	Percentage Difference
Pasteurized milk 1	0.25	1	300
Rice	0.52	3.26	526
Refined sugar	0.3	1.74	480
Raw sugar	0.17	0.87	412
Onionsz	0.54	2.6	381
Vegetable oil	0.87	10.53	1,110
Beans	0.46	5.28	1,048
Bananas, green	0.24	0.74	208
Bananas	0.22	0.4	82
Source: "Estudio Económico de America Latina y el Caribe, 1988, Cuba," Comisión Económica para America Latina y el Caribe, Naciones Unidas (Agosto 1989). 1Pesos per 946 gram jar.

Returning to Figure 1, a policy of partial liberalization would bring the price level from 1 to, say, 2, reducing the shortage to s2. This would appear to indicate a trade off between the size of the price increases and the magnitude of the shortage. But this trade off exists only as long as the excess demand curve remains at ED1. If the curve shifts up to ED2, for example, the shortage will increase to s2′. Thus, a policy of partial price liberalization coupled with loose control of aggregate demand (for example along the path A-B-C) would be associated with growing shortages and rising prices.

These considerations strengthen the case for rapid and full price liberalization. The experience of Eastern Europe demonstrates that such liberalization is possible. In Poland, for example, the share of sales at market-determined prices in the total value of sales increased from 50 percent in 1989 to 90 percent in early 1990. In the Czech and Slovak Republic, price liberalization in early 1991 brought the share of sales at free market prices to 85 percent of total sales. More recently, Bulgaria and Rumania have introduced comprehensive programs of price liberalization. In all four countries, however, governments have retained the power of intervention to limit “excessive” price increases resulting from monopolistic price setting.

The experience of Czechoslovakia is particularly illustrative in discussing the case of a country like Cuba where price distortions are pervasive. ¹¹ It is also particularly encouraging as it indicates that a comprehensive price reform need not evolve into a lasting inflationary process. The reform program launched in January 1991, which involved large-scale liberalization of both prices and foreign trade as well as a switch to world prices and convertible currencies in trade with members of the CMEA, did result in a sharp increase in prices: consumer and producer prices surged by 41 percent and 48 percent, respectively, in the first quarter of 1991. ¹² However, this was followed by a quick and pronounced deceleration of prices: by April-June 1991 the monthly rate of increase was down to an average of about 2 percent for consumer prices and less than 11/2 percent for producer prices. In recent months, prices in Czechoslovakia have been essentially stable.

The success in avoiding an inflationary spiral following price liberalization in Czechoslovakia was due to the pursuit of restrictive monetary and fiscal policies, including a swing from deficit to surplus in the government’s overall budget balance. It was also supported by measures to avoid excessive wage increases. In January 1991, agreement was reached to place a cap of 10 percent on wage increases during the year ended December 1991, with increases above this norm being penalized by prohibitive taxes on enterprises.

3. Price liberalization and the monetary overhang

By reducing the real value of the money supply, the one-shot increase in the price level associated with liberalization also would help to deal with any monetary overhang that may have built up over three decades of price controls and rationing. The monetary overhang, which is reported to be sizable in a number of Eastern European countries and particularly in the U.S.S.R., reflects the involuntary accumulation of financial saving associated with a prolonged excess demand for goods resulting from disequilibrium prices. A rise in the price level reduces the monetary overhang by acting as a tax on the real value of financial wealth.

An alternative way to reduce the overhang would be to cut the nominal value of the broad money supply through a monetary reform involving the write down or conversion of the financial balances held by the public. ¹³ This method, which appears to have worked well in several European countries in the period following World War II, does require difficult estimates of the size of the monetary overhang. Moreover, a monetary reform could raise political difficulties for a new government that is striving to gain the confidence of the population and establish its credibility. ¹⁴

No estimate of the monetary overhang is currently available in the case of Cuba. However, on the basis of data provided by the Economic Commission for Latin America, the monetary savings of the Cuban population increased by more than 240 percent from 1980 to 1988, compared to a rise of roughly 50 percent in the nominal value of Social Global Product. While these data are subject to a very wide margin of uncertainty–particularly as regards the Social Global Product–the implied magnitude of the fall in velocity suggests that the size of the cumulative excess demand for money could be quite large.

III. Trade Policy and the Balance of Payments

Felipe Pazos correctly emphasizes the issue of how and where to sell Cuba’s sugar production as one of the key problems of the transitional period. He may well be right in warning that, for an extended period, Cuba will have to continue selling an appreciable part of its crop to the U.S.S.R.–hopefully at world prices and in hard currency. Unfortunately, Soviet demand for imported sugar is unlikely to be sustained as domestic sugar production is rising. Therefore a critical item in Cuba’s foreign economic agenda will be the urgent need for negotiations aimed at the reestablishment of a sugar quota for Cuba in the U.S. market–unless, of course, the protectionist barriers and distortions that presently characterize the sugar sector in most industrial countries, notably in the European Community and the United States, were to be dismantled as a result of multilateral trade negotiations.

As important as it is, sugar will be only a part of Cuba’s foreign trade. Cuba’s external sector will cover a wide range of exportable and importable goods and services some of which, like tourism, are likely to grow much more rapidly than sugar exports during the transition. Thus, there is a general issue of what trade system would be most appropriate for Cuba. In this regard, Pazos observes that “after normal conditions are re-established, we should adopt a tariff regime with very low duty rates.” I fully agree, and I would add that tariff reduction should be accompanied–or even preceded–by the elimination of quantitative barriers on trade. However, Pazos goes on to caution that “the establishment of such a regime prior to the re-equipment of our industry and the re-establishment of our population’s will to work would result in a flood of imports and a serious disequilibrium in our balance of payments.”

Pazos is undoubtedly right that Cuba’s balance of payments is likely to come under pressure following the liberalization of its trade system. However, a number of forces will be set in motion that will tend to restore equilibrium in the external accounts. First, as Pazos notes, low levels of real incomes are likely to prevail at the start of the transition period, and this would tend to limit the demand for imports. Second, an incipient balance of payments’ deficit would result in a fall in the value of the peso-spontaneously under a floating exchange rate regime or through a discrete devaluation under a fixed but adjustable rate system–which would help to further reduce import demand and to stimulate exports and tourism (the implications of such a depreciation are examined in more detail in Section IV). Third, the adjustment process could be smoothed by obtaining temporary external financing, for example under an adjustment program with the I.M.F.; such a program would make particular sense in the context of a domestic plan to dismantle restrictions on foreign trade.

Balance of payments adjustment along the lines described above would appear to be far preferable to the maintenance of a system of import restrictions aimed at protecting the balance of payments. First, if freetrade is the ultimate objective–a point on which Pazos and I fully agree–a restrictive trade system would not avoid, but would only postpone, the need for balance of payments adjustment. Second, import protection, even if it is meant to be temporary, would run the risk of encouraging the organization of pressure groups with a vested interest in making protection permanent, and such groups may be difficult to reign in once they have been given the opportunity to flourish. This is a serious risk. Cuba should not replace the inefficiencies resulting from bureaucratic control of central planners with the inefficiencies resulting from an uncompetitive private sector operating under the protection of trade barriers.

It should be noted that rapid, large-scale trade liberalization is not merely an ideal. It has increasingly become a reality in many developing countries, including notably in Mexico, and in the majority of the formerly centrally planned economies. All the Eastern European countries have now abolished the state monopoly of foreign trade. And whereas various degrees of tariff protection remain in place, considerable progress has been made in reducing or eliminating quantitative restrictions. Up to 90 percent of all non-tariff import restrictions have been removed in Hungary and in the Czech and Slovak Republic and the proportion is even larger in Poland. Even in the Soviet Union, there has been a significant, albeit incomplete, reduction of government involvement in foreign trade.

IV. Exchange Rate Policy

Few decisions will be more important than the determination of Cuba’s exchange rate system. That decision will not only have a major influence on the evolution of Cuba’s balance of payments, it will also determine to a large extent the context in which the country’s monetary policy will operate. I will suggest that this is an area in which Pazos’ “separation principle” between what is appropriate for the transition and what is appropriate in the longer-run may have some relevance. Specifically, I will argue that while the best solution for the long term will be to establish a fixed parity between the Cuban peso and the U.S. dollar, the appropriate solution for the initial phase of the transition may need to involve a degree of exchange rate flexibility.

1. Long-term objective: fixed nominal exchange rate with the S. dollar

For the long term, the potential importance of the United States in the geographic structure of Cuba’s trade in goods and services, as well as in capital movements, suggests that a fixed nominal exchange rate with the U.S. dollar is likely to be the best solution. A fixed parity between the peso and the U.S. dollar would introduce an element of predictability for the bulk of Cuba’s external transactions. In particular, it would greatly facilitate transactions by U.S. tourists by allowing the U.S. dollar to circulate freely in Cuban territory. This kind of system worked well in Cuba prior to 1959 and it has worked very well in tourist-oriented economies such as the Bahamas and the member countries of the Eastern Caribbean Central Bank. ¹⁵

More importantly, a fixed exchange rate would provide an anchor for price stability as it would, in effect, tie the monetary policy of the Cuban central bank to that of the U.S. Federal Reserve. So long as the Federal Reserve continues to pursue a monetary policy aimed at resisting inflation in the short run, and at achieving a reasonable degree of price stability in the long run, a fixed exchange rate policy will ensure that Cuba’s price performance is also favorable. Cuba would thus avoid the episodes of high inflation (and the subsequent episodes of sharp monetary tightening that high inflation inevitably requires at some point) that have inflicted so much damage on the economies of many Latin American countries.

Another argument in favor of the fixed exchange rate system is that it would make it easier to protect the independence of Cuba’s central bank from eventual pressures from the government and its spending agencies. Of course, the exchange system will not guarantee independence as governments will always have the option of modifying the exchange rate rule and breaking the parity with the dollar. The avoidance of inflation will therefore require legislative guarantees including strict limitations on the provision of domestic credit. ¹⁶ Most importantly, it will require a sound fiscal policy that avoids the need–and therefore the temptation–to resort to the inflation tax. Of course, this will require firm control of public expenditures and a strong system of revenue collection. It may also require legal restrictions on the size and on the rate of expansion of the public debt.

It could be argued that a fixed exchange rate system would imply a loss of independence of domestic monetary policy and thus would deprive the Cuban authorities from a powerful tool of demand management. For several reasons, I find this argument to be unconvincing.

First, the ability of monetary policy to influence the real economy should not be exaggerated. In the long run, monetary policy cannot influence either the real interest rate or the real exchange rate. Accordingly, monetary policy cannot, beyond the short run, influence real economic magnitudes such as output and employment. As emphasized in Section I, these variables will be determined by the factors that account for the long-run growth of productive capacity: capital formation and saving; the intensity and the quality of the labor force; and total factor productivity–which means technological progress and the whole spectrum of managerial, administrative and structural factors that influence the overall level of economic efficiency.

There is, however, one way in which monetary policy can influence investment and productivity in the long run. It can do so by helping to establish a framework of macroeconomic stability and confidence in which the private sector, both domestic and foreign, can operate at its best. First and foremost, monetary policy can ensure that inflation remains at levels sufficiently low that they will not hinder investment and economic activity by distorting resource allocation and investment decisions, by creating uncertainty about future prices, wages and interest rates, and by encouraging activities whose sole purpose is to hedge against inflation. Therefore, if the “loss of independence” of monetary policy means the maintenance of price stability, this loss will be a blessing in disguise.

Second, it is not realistic to count on monetary policy to isolate a small economy like Cuba from cyclical fluctuations in its major trading partners. For example, in the event of a recession in the United States, Cuba’s economy unavoidably would be affected through a reduced demand for Cuban goods and services by U.S. importers and tourists. However, the intensity of these adverse effects would be cushioned by the operation of automatic stabilizers, both in the United States and in Cuba. For example, a fall in economic activity in the United States would exert downward pressure on U.S. interest rates and on the value of the U.S. dollar. Under a fixed exchange rate system, Cuban interest rates also would be allowed to fall, thus boosting the interest-sensitive components of domestic demand, and the Peso would depreciate against other major currencies, thus stimulating Cuban exports to countries other than the United States. Exports to the United States also would rise, as the competitive position of Cuban producers in the U.S. market would improve owing to the Peso’s depreciation vis-à-vis third currencies.

Automatic stabilizers also would operate in response to fluctuations in domestic output. For example, a cyclical decline in domestic economic activity would temporarily reduce tax revenue and thus cushion the decline in disposable income; unemployment compensation payments also would provide a stabilizing influence. It is true, however, that under a fixed exchange rate system the country would forego the additional stabilizing effect that could result from a depreciation of the Peso against the U.S. dollar and from a cyclical decline in domestic interest rates.

The need for an instrument of domestic demand management may arise in particular in the event of a major but temporary change in the terms of trade, for example as a result of a large drop in the world price of sugar. ¹⁷ In that situation, there is a question whether the stabilizing role of fiscal policy should be limited to the working of automatic mechanisms or whether within limits, there may be a role for an active fiscal policy as a tool of stabilization. The limits are likely to be narrow. Longer-term objectives require that expenditure be kept under firm control, and the basic structure of the tax system clearly cannot be altered frequently without introducing serious costs in terms of uncertainty. Moreover, the free mobility of goods, capital, skilled labor and entrepreneurship will mean that tax arbitrage will set a limit to the extent to which taxes in Cuba can differ from those in the United States. Nevertheless, there could be scope for using temporary tax cuts and the front-loading of certain expenditure projects to provide temporary support for aggregate demand and economic activity in exceptional circumstances. The challenge will be to introduce the appropriate rules and mechanisms to ensure that these temporary fiscal measures do not give rise to lasting departures from the long-term budgetary goals.

2. Exchange rate policy in the transition

So much for the long term. The next question is which exchange rate system is feasible and desirable for the transitional period. I will consider two possibilities: (i) a fixed nominal exchange rate vis-à-vis the

U.S. dollar and (ii) a temporary floating exchange rate system. I will discard a third possibility: exchange controls, as it would involve serious distortions and would be inconsistent with other elements of the transitional strategy, including in particular full and rapid trade liberalization.

a. Fixed nominal exchange rate

The first possibility would be to fix the value of the peso at an early stage, probably following a large, once-and-for-all devaluation. It could be argued that the early establishment of a fixed nominal exchange rate vis-à-vis the U.S. dollar would have all the advantages mentioned in the previous section’s discussion of exchange rate policy in the longer term. In particular, it would provide an “anchor” to a noninflationary monetary policy and help to bring about a rapid convergence of domestic and U.S. inflation. If monetary policy were, for some time, more expansionary than that compatible with the maintenance of the fixed exchange rate, the central bank would lose international reserves, and it would be forced to reduce the growth of money and credit. If the authorities’ commitment to the fixed rate system is credible, the convergence of inflation would take place without significant costs in terms of output and employment.

Several objections could be raised, however. ¹⁸ First and foremost is the issue of credibility. If the authorities’ commitment to the fixed exchange rate is not credible, central bank reserves, which initially are likely to be quite small, could come under speculative attack, ultimately forcing the abandonment of the system and the adoption of a floating rate system and/or exchange controls. Even if the central bank can hold the line for some time serious problems could arise. If monetary policy is sufficiently tight but inflationary expectations fail to adjust–because the public expects that monetary restraint will not be maintained–domestic inflation will continue to outpace external inflation and the real exchange rate will appreciate. This would have adverse effects on exports, output and employment; the current account position will deteriorate, thus further eroding confidence in the authorities’ ability to maintain the exchange rate. Perseverance with a tight monetary policy eventually will convince markets that the exchange rate will be maintained, and the real exchange rate will gradually depreciate. But the process of stabilization without credibility could involve serious output losses.

Second, there may be serious difficulties in establishing the exchange value of the peso in the initial stage of the transition. If the initial exchange rate is seriously overvalued, there could be an immediate speculative attack on the central bank. If reserves are inadequate, this could force the abandonment of the currency peg, perhaps under disorderly conditions, with possibly lasting adverse consequences for the credibility of a nominal exchange rate anchor. The early establishment of a fixed exchange rate will be complicated by the fact that price liberalization most likely would result in a radical change in the structure of relative prices and costs. Because the present structure of relative prices and costs is so distorted, it is virtually impossible to establish a priori the level of the exchange rate that would ensure, in the short term, an appropriate level of competitiveness and satisfactory balance in the country’s external transactions.

b. Floating exchange rate

In view of the fundamental uncertainty as to how the cost structure will evolve, it could be argued that a transitional system of floating exchange rate would have considerable advantages. In particular, it would allow the exchange rate to “find its value” in the foreign exchange market as trade flows adapt to the country’s cost/price structure while avoiding the risk of speculative crisis that would be inherent in a process of occasional, discrete adjustments in temporarily fixed exchange rates.

Two important questions arise about the implications of this initial period of floating exchange rate. The first question relates to the strong possibility that the introduction of external convertibility and the creation of a free foreign exchange market would result in a sharp fall in the value of the peso. In this regard, Pazos notes that “if the disequilibria of the transition were to result in a strong depreciation of the peso, and consequently in a fall in the foreign currency value of all Cuban assets, the free entry of foreign capital would be tantamount to giving away Cuba’s wealth to the new investors.” ¹⁹

It is not clear, how such a depreciation could be avoided. One possible interpretation of Pazo’s concern is that a system of temporary controls on capital inflows would be appropriate. However, such a system would deprive Cuba of a source of external capital at a time when such capital would be most urgently needed and most reluctant to flow in. Furthermore, capital controls would deprive the country from an automatic mechanism of balance of payments adjustment. By lowering the foreign currency price of domestic assets, a devaluation of the peso would increase the expected (risk-adjusted) rate of return on such assets from a level which, initially, is likely to be very low. The resulting inflow of foreign capital would, for some time, finance a higher current account deficit than would otherwise be possible, thus allowing a higher level of capital goods’ imports and giving time for the country’s export capacity to develop.

The second question relates to the conduct of monetary policy. Without the nominal anchor provided by a fixed exchange rate, monetary policy would have to target on some monetary or credit aggregate. But monetary targeting, often a technically difficult matter, is likely to be particularly complicated in the wake of price liberalization, when monetary velocity is likely to experience large changes. In those circumstances, the objectives of monetary policy will be difficult to explain and its results hard to evaluate. Without the yardstick provided by the nominal exchange rate, monetary policy would be unable to establish credibility and would therefore fail to keep inflation under control. This is a serious risk. But it is a risk that argues against a floating rate system as a permanent or even a lasting solution, and does not rule out that such a system might play a useful role for a relatively brief transitional period, particularly if care is taken to avoid the (formal or informal) indexation of nominal wages.

c. A possible solution

The observations made above suggest that exchange rate policy might go through three distinct phases. In the first phase, a program of price liberalization and subsidy reduction will be designed and formally announced, and the exchange rate will be allowed to float. Most probably, as suggested by the analysis of Calvo and Frenkel, the Peso will then depreciate in anticipation of price reform. ²⁰ A number of other important things would be done during this phase. First, legislation guaranteeing the independence of the central bank and imposing strict limitations on the expansion of domestic assets would be enacted.

Second, a stabilization fund could be negotiated with the United States and other industrial countries and, if possible, with support from international financial institutions. Third, trade and price liberalization will be implemented which, most probably, would result in a further depreciation of the peso. This would be accompanied by restrictive monetary and fiscal policies with a view to bringing down the rate of price increase to low levels over a relatively short period, following the recent example of Czechoslovakia. In the second phase, a crawling peg would be introduced, with a pre-announced schedule of progressively smaller devaluations that would continue until convergence between Cuban and U.S. inflation is achieved. The third phase would then begin with the establishment of a fixed nominal exchange rate against the dollar.