Fiscal Proxies for Devaluation: A General Review
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

JOHN F. LAKER *

Abstract

JOHN F. LAKER *

JOHN F. LAKER *

The question of using the tax system to substitute for exchange rate changes is an old and controversial one. As early as 1931, Keynes advocated a system of import tariffs and export bounties for commodity trade as an alternative to devaluation of sterling at that time. Keynes and others—Great Britain (1931, p. 199)—argued that such a system would be superior because it would avoid the depreciation—in terms of gold—of British foreign obligations and assets denominated in sterling. 1 This argument has its echoes in the current problems faced by developing countries with external debt denominated in hard currencies. In contrast, Hicks’s (1959) proposal for a similar fiscal proxy for devaluation had a quite different motivation. Hicks saw this type of approach as the way to avoid the nationalistic terms-of-trade exploitation of other countries, resulting from the use of import restrictions rather than devaluation to overcome balance of payments or unemployment problems.

The main feature of such tax/subsidy proposals is that they are restricted to transactions in the goods market. Their underlying rationale is the long-established proposition, derived from partial-equilibrium analysis, that the effects of a devaluation on the goods market can be duplicated exactly by an explicit tax at the same (uniform) rate on imports and, simultaneously, a subsidy at the same (uniform) rate on exports.2 This notion of uniformity in tax and subsidy rates was not, however, always a feature of tax/subsidy proposals: Keynes, for example, later suggested that different percentage taxes and subsidies should apply to different commodities.

Since 1945 a number of countries, in various stages of development, have used fiscal measures to forestall an explicit exchange rate change. This usage has been marked by the wide variety of techniques adopted and by the frequency of departures from the uniformity principle. In general, the lessons from this experience have been discouraging. As Cooper (1973, p. 170) noted, a common problem for developing countries has been that “when balance-of-payments pressures develop (sometimes as a result of inflationary policies, which in the short run are often also a successfully ambiguous way to reconcile conflicting social objectives), officials engage in a series of patchwork efforts and marginal adjustments to make the problem go away (raising tariffs here, prohibiting payments there), which may disturb the original objectives as well as coping only inadequately with the payments difficulty.”

Recent developments suggest that another look at the role of fiscal proxies may be fruitful. On the theoretical level, recent “portfolio balance” models of the balance of payments, which go beyond the standard goods-market-only analysis, have provided new insights into the short-run and long-run effects of exchange rate changes, raising the question whether the simple equivalence between devaluation and a uniform tax/subsidy scheme holds within a general-equilibrium rather than a partial-equilibrium framework. On a practical level, despite the greater flexibility in the use of the exchange rate mechanism in the past decade, many countries in balance of payments difficulties still regard devaluation as a measure of last resort and seek a variety of partial substitutes before devaluation is finally and reluctantly undertaken. Under these circumstances, the possibility of securing the important effects of devaluation by fiscal proxies, without the inefficiencies and distortions of earlier “patchwork” approaches, is a highly relevant question.

This review of fiscal proxies for devaluation addresses these theoretical and practical questions. Section I summarizes the major effects of exchange rate changes as emphasized in standard balance of payments theory and in recent general-equilibrium approaches, and analyzes the extent to which these effects could, in theory, be duplicated by manipulating the tax system. Section II examines the practical difficulties involved in establishing an equivalent tax regime, drawing on the experience of both developed and developing countries. The final section presents the main conclusions of the review.

Throughout the paper, the analysis of fiscal proxies proceeds on the assumption that a change in the exchange rate—however secured—is the appropriate policy response to disequilibrium in the balance of payments; a different set of questions revolving around the choice of devaluation or deflationary fiscal and monetary policies is not discussed.

I. Theoretical Analysis of Devaluation and Fiscal Proxies

SHORT-RUN AND LONG-RUN EFFECTS OF DEVALUATION

The conventional view of the effects of an exchange rate change is generally grounded in a Keynesian model of income determination, in which the exchange rate determines the relative prices of traded goods and, hence, international competitiveness. This so-called standard model 3 focuses on the balance of trade (i.e., the goods market), on the assumption that the effects of an exchange rate change on this balance would dominate any effects on the other accounts of the balance of payments. The model also assumes, crucially, that the monetary consequences of any change in the trade balance can be and are absorbed (sterilized) by the monetary authorities.

In this standard model, devaluation has both price-adjustment and income-adjustment effects. The former are treated as impact effects and are analyzed in terms of the elasticities approach. 4 Within this partial-equilibrium framework, it is easy to demonstrate that devaluation is exactly equivalent to a uniform ad valorem tax on imports imposed simultaneously with a uniform ad valorem subsidy to exports. Throughout the paper, this particular proxy is defined, for simplicity, as a uniform commercial policy. 5 The income-adjustment effects refer to the changes in aggregate incomes induced by the initial change in the trade balance. Assuming a positive marginal propensity to save, any impact improvement in the trade balance will generate a multiplier increase in income that will offset this improvement, although only partially. Under these conditions, the standard model concludes that devaluation will lead to a permanent improvement in the balance of payments. Whether a uniform commercial policy produces the same result is less straightforward. If this policy led initially to an improvement in the trade balance, its budgetary cost would be positive, involving a net injection by the government to aggregate demand. This action would create a need for financing, which could be met by expanding the money supply, by the sale of government assets, or by the issue of government debt. All these methods, however, involve a cumulative disturbance to portfolio balance, the implications of which cannot be traced out in the framework of the standard model.

The difficulty in analyzing the monetary implications of the uniform commercial policy reflects a major weakness in the standard theory: although based on a macroeconomic model, the theory is not truly a general-equilibrium one, since it ignores the interactions between the goods market and the money market. In contrast, the recent portfolio balance approach to balance of payments theory places its major emphasis on these interactions. 6 This approach assumes that monetary inflows or outflows associated with balance of payments surpluses or deficits are not sterilized—or cannot be, within a period relevant to policy analysis—but instead influence the domestic money supply. It further assumes that the relationship between the demand for and the supply of money (in simple formulations) or wealth (in more elaborate models) plays a major role in the functioning of all markets in the economy.

The analysis of devaluation under the portfolio balance approach is well known and can be summarized using the skeletal general-equilibrium model introduced by Kemp (1970) and Mundell (1971, Ch. 9). It is a small-country, full-employment model with only one asset market—that for money balances. In this framework, external trade is viewed as the exchange of goods for money, so as to achieve stock equilibrium in the money market; the role of the exchange rate is to define the price of money in terms of goods. The impact effect of devaluation is an increase in the domestic currency price of traded goods, producing a reduction in the real value of cash balances. This creates a stock excess demand for money and causes expenditure to decline relative to income as residents “hoard” in order to restore real balances. If the domestic source component of the money supply is kept constant, the increase in the nominal quantity of money demanded must be met by an inflow of foreign exchange, via a cumulative trade balance surplus. Thus, if the market for money balances is assumed to take time to adjust, devaluation will cause a temporary improvement in the balance of payments, during the period of transition to the new equilibrium. The trade balance surplus represents a phase of stock adjustment in the money market, but it cannot be regarded, as it is in the standard model, as a permanent (equilibrium) flow.

In the long run, when real balances are restored to desired levels, the model predicts that devaluation has no permanent effect on the balance of payments. Its only long-run effects are an increase in the nominal quantity of money (in the form of higher international reserves) and an increase in nominal prices of both traded and nontraded goods proportional to the rate of devaluation. This conclusion also holds in more elaborate models that include other asset markets besides the market for money balances. Under certain circumstances, nevertheless, devaluation may be a useful policy tool: it may be used to achieve a once-for-all increase in a country’s international reserves, or to finance a temporary budget deficit through money creation, without causing a temporary deterioration in the balance of payments. In the latter case, devaluation would reduce private real expenditure and thus release resources for the public sector, while the attendant increase in prices would encourage residents to add to their cash holdings the money issued to finance the budget deficit. In more realistic models allowing for price and wage rigidity, devaluation also plays a role when balance of payments disequilibrium is due to real wages being too high for the traded goods sector to be competitive. Under these circumstances, devaluation could serve to reduce real wages by increasing the domestic price level, and this policy would be superior to deflation in enabling full employment to be reached (Dornbusch (1974)).

SHORT-RUN AND LONG-RUN EFFECTS OF FISCAL PROXIES

In the portfolio balance approach, the driving force of devaluation comes from the proportional increase in the price of traded goods, which prompts the expenditure adjustments necessary to restore real cash holdings. An exogenous proportional increase in these prices can also, of course, be secured by a variety of tax measures, and the conditions under which these measures are equivalent to devaluation have been established—within the Kemp-Mundell framework—by Berglas (1974) and Kuska (1976). Two types of fiscal proxy are discussed in turn.

Uniform commercial policy

The traditional alternative to devaluation is the uniform commercial policy, defined earlier as an equiproportionate subsidy on all exports, combined with a tariff at the same rate on all imports. The initial impact of this policy is an increase in prices of traded goods equal to that achieved by devaluation. For this reason, the short-term adjustment process, working through the decline in real money balances and aggregate expenditure, must be the same for both alternatives, as is the long-run equilibrium result that there is no permanent improvement in the balance of payments. Whether the uniform commercial policy has an identical impact on the net accumulation of foreign exchange, however, will depend on how the policy is financed. Any initial improvement in the trade balance (starting from an equilibrium position) implies that the net budgetary outlay of the fiscal proxy is positive, and this could be financed by monetary means, that is, by newly printed money. Under these conditions, only part of the initial decline in real money balances is made up by the inflow of foreign exchange; the remainder is provided by the government through the financing of the uniform commercial policy. Hence, the balance of payments during the adjustment process and the net accumulation of foreign exchange in long-run equilibrium, when expressed in domestic currency, will be proportionately lower under the fiscal proxy than under devaluation. This conclusion, which is demonstrated formally in the Appendix, qualifies the simple equivalence of the two policies established in the goods-market-only framework. However, when expressed in terms of foreign currency, both the balance of payments and the final foreign exchange position will be identical under the two policies. Alternatively, the net budgetary outlay of the fiscal proxy could be covered by a general tax adjustment of equal yield, so that residents would be forced to adjust their real money balances entirely through the balance of payments, as with devaluation. If so, both policies would have the same balance of payments impact measured in domestic currency; the fiscal proxy, however, would produce a larger change in foreign exchange reserves when expressed in terms of foreign currency. 7

Selective taxes

Selective taxes represent a second fiscal alternative to an exchange rate change. Consider first a flat-rate production (consumption) tax on nontraded goods, the proceeds of which are completely distributed as transfer payments. Assuming no income redistribution effects, the underlying demand functions in the economy will remain unchanged after the tax, but the supply functions must shift, since the tax drives a wedge between producer and consumer prices. Its initial impact is, therefore, a reduction in the supply of nontraded goods, as producers shift resources into production of traded goods in response to the relatively higher producer price of the latter. The price of nontraded goods must rise to eliminate this excess demand, leading in the short run to an excess demand for money and an excess supply of traded commodities. The resultant balance of payments surplus produces an increase in the quantity of money, which continues until a new equilibrium is established at a higher ratio of nontraded to traded goods prices. Thus, the (temporary) balance of payments effect of a production tax on nontraded goods is similar to that of devaluation, but the long-run production and consumption pattern resulting from the tax is different. Long-run market clearing is associated with an increase in the quantity of traded goods demanded, requiring a movement in resources from the nontraded goods sector. Since the price of traded goods is given, this movement must be induced by a decline in the producer price of the nontraded good. The long-run result, therefore, is a rise in the price of the nontraded good by less than the tax rate, and a shift in both production and consumption in favor of traded goods.

The analysis of a production tax on traded goods is similar. Given the small-country assumption, the tax will not affect the price of traded goods, but it will induce producers to shift out of their production; hence, its introduction will mean an excess supply of nontraded goods and a decline in their price. The result is an excess supply of money and an excess demand for traded goods, producing a balance of payments deficit that will continue until the economy converges to a long-run equilibrium position. The long-run effect of the tax is a decline in the price of the nontraded good by less than the tax rate, and a reduction in the production and consumption of traded goods.

The production tax on traded goods could be supplemented by a tax at the same rate on imports, and an equivalent rebate for exports. With this border-tax adjustment (on the destination principle), the tax system is equivalent to a consumption tax on traded commodities, or to a production tax accompanied by a devaluation of the same rate. Under this system, there are two offsetting short-run effects on the balance of payments: a production tax effect resulting in a balance of payments deficit, and a devaluation effect leading to a balance of payments surplus. The net outcome of these effects depends on the explicit form of the demand and supply equations; in any case, the balance of payments deficit will be smaller once the border-tax adjustment is introduced. The border-tax adjustment will not, however, affect long-run relative prices or the pattern of production or consumption established by the production tax.

To complete the theoretical analysis of fiscal proxies, one basic limitation of the Kemp-Mundell framework must be recognized. Despite its general equilibrium outlook, the simplifying assumption of only one asset (money) rules out detailed consideration of asset market behavior, and this may be a serious omission for economies with highly developed and integrated systems of financial intermediation. More elaborate portfolio balance models have addressed this problem by broadening the spectrum of financial assets to include bonds and other assets, and examining the role the exchange rate plays in balancing stock demands for and supplies of these assets. 8 These models confirm the long-run results of the Kemp-Mundell framework: devaluation causes all nominal variables (including asset holdings) to rise by the same amount as the exchange rate, but all real variables remain unchanged. The short-run impact of devaluation may, however, be different, because of the possibility of capital gains or losses on financial assets attributable to the devaluation. For example, a country in a net creditor position in assets denominated in foreign currency enjoys an increase in nominal wealth from devaluation. A reduction in its real wealth—which triggers the expenditure adjustments central to the portfolio balance approach—comes only if the effect of devaluation on the general price level is stronger. If it is not, the resultant increase in real wealth would reinforce the excess demand for nontraded goods resulting from devaluation, thereby reducing the extent of adjustment necessary to restore the original relative price ratio between traded and nontraded goods. Capital gains or losses thus become an important element in determining the amount of “hoarding” needed to restore long-run equilibrium and, hence, in determining the maximum (temporary) impact that devaluation may have on the balance of payments.

This focus on portfolio adjustments raises the question of whether fiscal proxies may have asset market repercussions that are different from devaluation, further qualifying the traditional equivalence between the two policies. On this matter, two basic conclusions can be derived from the more elaborate portfolio balance models. First, the price and wealth effects of devaluation can be reproduced exactly only by a tax/subsidy scheme that imposes an equivalent tax on all purchases of foreign exchange from the monetary authority, combined with an equivalent subsidy on all sales of foreign exchange. Only by covering all foreign exchange transactions can a fiscal proxy duplicate any initial portfolio adjustments in response to the higher price of traded goods, and the short-run expenditure effects resulting from the change in both nominal wealth (through capital gains or losses) and real wealth. Again, however, the financing of the fiscal intervention will determine whether its balance of payments impact differs from that of devaluation. If the scheme is financed by new money or the issue of government bonds to residents, the final change in foreign exchange reserves will be smaller (measured in terms of domestic but not foreign currency), since part of any initial fall in real wealth is made up by the assets created to finance the scheme. An identical increase in foreign exchange reserves in domestic-currency terms requires a financing method (such as a nondistorting general tax change) that forces any change in real wealth to be made up entirely through the balance of payments.

Second, and as a corollary, a uniform commercial policy that applies the tax/subsidy arrangement only to goods market transactions must involve a different set of short-run price and wealth effects and, hence, a different adjustment process. The narrower coverage rules out capital gains or losses on instruments denominated in foreign currency (and changes in the real burden of debt similarly denominated). Hence, there are no nominal wealth changes from the scheme itself to reinforce, or to offset, the short-run increase in the prices of nontraded goods and the resultant decline in real wealth. The required adjustment in real wealth will therefore differ from that associated with devaluation, depending on the size and initial currency composition of portfolios, but the long-run real outcome of the two policies is the same, at least in fixed-factor models.

In dynamic versions of the portfolio balance approach, which allow for a growing capital stock, a further qualification to the traditional equivalence is required when the uniform tax/subsidy scheme is applied to all current account transactions. Such a scheme corresponds to a dual exchange rate system in its pure form, in which all capital items are transacted at the official exchange rate, but all transactions for goods and invisibles—including interest and profits—take place at a depreciated rate by virtue of the tax/subsidy arrangements. This premium on the official exchange rate reduces the effective rate of return on bonds denominated in domestic currency for nonresidents, but raises the rate on such bonds denominated in foreign currency for residents. A country in a net creditor position in these latter assets therefore enjoys an increase in its nominal wealth following introduction of the fiscal proxy, so that the scheme duplicates the short-run capital-gains effects of devaluation. In the long run, however, the maintenance of a differential in effective exchange rates involves a permanent subsidy on lending abroad, and this will have implications for both the currency composition of portfolios and the equilibrium stock of capital, which are absent from a uniform exchange rate change. In general, the long-run effects of such a subsidy are ambiguous; as Dornbusch (1975 b) has shown, divergences from the devaluation result can be spelled out only in dynamic portfolio balance models in which asset demand functions are given specific form.

In summary, the portfolio balance approach has established more rigorously the conditions under which the important effects of devaluation can be proxied through the tax system. It has confirmed that an identical short-run and (neutral) long-run impact on resource allocation can indeed be secured by a tax/ subsidy scheme, but only if it is applied to all foreign exchange transactions. Whether such a scheme has the same impact on the balance of payments outturn (in domestic-currency terms), however, depends on how it is financed. The ability to analyze the financing issue explicitly distinguishes the portfolio balance approach from the earlier partial analyses of fiscal proxies. The approach has also shown that alternative tax/subsidy schemes—applied either to a particular subset of foreign exchange transactions or selectively to traded or nontraded goods—may also have a positive impact on the balance of payments in the short run but, unlike devaluation, may result in long-run changes in the composition of production and consumption. Of course, the practical policy relevance of the portfolio balance approach sketched in this paper has often been questioned. 9 Its basic assumptions have been viewed as overly narrow for many country situations, particularly for developing countries, while the strength of the real balance effect and the period during which it comes fully into play are not known with any certainty. 10 These issues are not addressed in this review, however, since neither modification of the approach nor the attempt to give it empirical flesh alters the basic conclusions on the relationship between devaluation and fiscal proxies.

II. Practical Limitations of Fiscal Proxies

The preceding section has shown that, in theory, the real effects of devaluation, and its impact on foreign exchange reserves, can be duplicated exactly by an appropriately financed tax/subsidy scheme, extended uniformly to all transactions. Actual experience with fiscal proxies, however, has shown that it is extremely difficult to establish this equivalence in practice. The major problems are that (1) the tax/subsidy scheme is often applied only to some subset of current account transactions; (2) the scheme is not always applied on a uniform basis, even to current account transactions; (3) the administrative costs of implementing the scheme may be significantly greater than those involved in a straightforward exchange rate change; and (4) the scheme may ultimately lead to widespread corruption, evasion, or misuse, particularly if the tax and subsidy rates rise considerably.

These difficulties can be illustrated by reference to a number of examples where the tax system was used specifically to duplicate the balance of payments impact of an exchange rate change, rather than (as with a particular tariff on certain commodities) to influence the long-run structure of production and trade. The episodes cover the experience of France (1957–58), Israel (1955–62), India (1963–66), and the Federal Republic of Germany (1968–69). This list is not exhaustive: for example, the complex tax/subsidy schemes used to compensate for overvalued exchange rates in a number of Latin American countries are not considered. Nevertheless, the four examples—drawn from countries in very different stages of development—are sufficient to illustrate the variety in fiscal techniques used and the problems involved in attempts to proxy exchange rate changes.

CREATION OF A DUAL EXCHANGE MARKET SYSTEM

Despite the theoretical requirement that a fiscal proxy cover all foreign exchange transactions, tax/subsidy schemes in practice have generally been confined to goods and a few transactions in invisibles (such as remittances and tourism). Unless there is a separate balance of payments objective with regard to the long-term capital account, these limitations result in the creation of a dual exchange market system without obvious economic rationale. As noted earlier, the long-run resource-allocation effects of such a system are ambiguous, but its short-term operations are well known, as is the conclusion that it is generally not administratively practicable to achieve full segregation of goods and asset market transactions to enable this type of system to operate effectively. 11 The likelihood of loopholes further weakens the correspondence between the (partial) fiscal proxy and devaluation. For example, if some flexibility is allowed in the timing of current payments, the impact of a tax applying only to goods may be deferred through the longer-than-usual receipt of trade credit, passing through at a premium at the official exchange rate. This may soften the impact of the fiscal proxy, by slowing the rise in the price of traded goods that triggers the switch in expenditure to nontraded goods.

Of the four examples under review, only the French experience comes close to satisfying the requirement of complete coverage. In August 1957, France effectively reduced the exchange rate of the franc by introducing a 20 per cent surcharge on external payments and an equivalent premium on external receipts. The scheme operated through the Exchange Stabilization Fund and applied to about 60 per cent of all imports (excluding coal, petroleum products, and major raw materials), about 65 per cent of all exports (excluding iron and steel products and textiles, the latter being covered later), and all other transactions.

In contrast, the fiscal proxies adopted by the three other countries were confined solely to trade in goods. The Federal Republic of Germany in November 1968 effectively revalued the deutsche mark for goods market transactions, by the use of border-tax adjustments. This procedure involved the introduction of special taxation, which took the form of a differentiated turnover tax on exports and an equivalent subsidy on imports not subject to the agricultural arrangements of the European Community. All other transactions continued to pass through at the (depreciated) official exchange rate, severe measures being introduced at the same time to deter the inflow of speculative funds. Israel maintained an unchanged par value from 1955 until February 1962, but frequent increases in tariff rates, special import levies, and export subsidies produced a devaluation over this period of about 21 per cent in the exchange rate applicable to commodity imports, and about 24 per cent for exports of goods and services. 12 In contrast to the previous examples, the Israeli system involved an explicit dual exchange rate system, under which many payments for invisibles and capital transactions were effected through a separate exchange market, known as the securities free market. By 1959 the tax on onward transfers through this market ranged from 30 per cent to 35 per cent. An outright premium of 20 per cent was paid on certain other inward personal transfers and on foreign currency exchanged by tourists. Finally, India used a complex system of export subsidies to compensate exporters for an overvalued exchange rate in the early 1960s, and in February 1965 introduced an additional ad valorem import duty of 10 per cent on all imports, excluding foodgrains, fertilizers, pesticides, crude petroleum, and, subsequently, raw jute.

BREAKDOWN OF THE UNIFORMITY PRINCIPLE

In many instances, fiscal proxies—even on goods—have not been applied uniformly, but on a selective and discriminatory basis. In such cases, the essential neutrality of an exchange rate change for resource allocation within the traded goods sector is lost, and the fiscal proxy can become the source of serious distortions in relative prices. Country experience suggests that this is the most telling objection to the use of the tax system for exchange rate adjustment.

The only fiscal proxy reviewed here that met the uniformity requirement was that introduced by France. In the Federal Republic of Germany, the special tax/subsidy scheme was applied at a rate of 4 per cent to most eligible goods but at half this rate for others. 13 However, the type of distortions attributable to fiscal proxies is most clearly illustrated by the examples of Israel and India. In Israel, the network of taxes and subsidies on imports and exports constituted an implicit multiple exchange rate system, characterized by its complexity and its discriminatory nature. On the import side, the system of tariffs and quota restrictions tended to discriminate against consumer goods, in favor of raw materials and investment goods. To maintain price stability, imports of food, agricultural raw materials, pharmaceuticals, and petroleum products were subsidized (until the end of 1958); capital goods, and raw materials used to produce export goods, were exempt from tariff duties; and luxury imports were subject to special levies in addition to the existing customs tariff. Special levies were also imposed on imports of raw materials as they were freed from quantitative restrictions, both to restrain the increased demand for these items and to compensate for the relatively appreciated exchange rate.

Export subsidies took the form of retention-quota privileges and premiums on sales of foreign exchange. For subsidization purposes, the proceeds of a given export transaction were divided into value replaced—the amount of foreign currency directly or indirectly required to produce the good—and value added—the net foreign currency gain realized by the sale of the export. Under the first arrangement, exporters of industrial products were permitted to keep part or all of their export proceeds in special nontransferable foreign currency accounts. The value-replaced component was retained as a self-renewing foreign currency fund, but the value-added component could be used to finance purchases of raw materials for direct sale on the local market or as inputs for goods produced for the local market. Since imports of raw materials at that time were effectively restricted by quota, this arrangement gave exporters an element of monopolistic power, through their right to import over and above existing quotas. The resultant windfall profits thus constituted an indirect form of export subsidization. After 1955 direct subsidies began to replace retention-quota privileges on industrial exports, and by the end of 1959 all retention quotas had been abolished.

Direct subsidies were first applied mainly to diamonds; they were extended to citrus and other agricultural exports from 1956, and by the end of 1959 applied to all commodity exports and exports of shipping and aviation services. By that time, the subsidy rate paid on value added had increased to 47 per cent for most exports; higher rates (up to 63 per cent) were allowed for exports to specific countries, or for marginal increases in export proceeds over a minimum level determined by past performance.

The outcome of these various fiscal interventions—together with different exchange rates applicable to various types of transactions in invisibles and capital—was a multiplicity of effective exchange rates. By the end of 1961, the distortive effects of this system on the domestic price structure had become a major policy concern for the authorities.

The attempt of India to avoid a formal exchange rate change in the early 1960s involved an even more elaborate and cumbersome network of taxes, licenses, and subsidies on foreign trade. 14 Virtually all imports required individual licenses, with foodstuffs, capital goods, industrial raw materials, and other essential commodities generally given priority. In addition, imports of capital goods were subject to special procedures, which involved a review of import license applications by a capital goods committee; these procedures allowed a considerable degree of administrative discretion in the granting of import licenses. Imports were also subject to a range of customs duties, augmented in February 1965 by an across-the-board surcharge of 10 per cent on most imports.

Export subsidization took the form of specific fiscal measures and import entitlement schemes similar, in principle, to Israel’s retention-quota scheme. Fiscal measures included exemptions from sales taxes on final sales, refunds of indirect taxes on inputs, direct tax concessions, outright subsidies, and rail freight concessions. From 1963 onward, however, the principal instruments of export promotion were the import entitlement schemes, under which eligible exporters received import licenses in proportion to the value of exports earned. Since the entitlements were generally transferable within a scheme, they could be sold for a considerable premium, particularly with the stringent import rationing in operation. In general, the import entitlement was to be determined on the uniform basis of twice the import content, subject to a maximum of 75 per cent of the f.o.b. value of exports. However, the schemes were also tailored to the requirements of specific industries, which led to considerable departures from the general rule. Between 1963 and 1965, for example, new export promotion schemes, covering almost half the export earnings eligible for this form of assistance, offered import entitlements in excess of 75 per cent and up to 100 per cent of f.o.b. value in certain cases. In addition, there were frequent changes in the export incentives offered by the schemes, because of administrative changes in their coverage and in entitlements for given products, and changes in premiums on the entitlements themselves. The net effect of these schemes was, therefore, a highly variable and discriminatory system of implicit export subsidies.

This myriad of administrative interventions and controls on foreign trade became a heavy weight on Indian commerce, and was responsible for one of the more complex restrictive systems. Moreover, the system was increasingly subject to ad hoc adjustments as balance of payments difficulties became more pronounced, which put even greater pressure on the network of administrative controls and further distorted the price structure.

ADMINISTRATIVE BURDEN OF FISCAL PROXIES

As shown by India’s experience, the administrative costs of implementing an exchange rate proxy through the fiscal authorities will be higher—significantly so in many cases—than for a straightforward adjustment in the exchange rate itself. Further, administrative difficulties may be the main reason why the proxy cannot be extended to all foreign exchange transactions, as the equivalence principle requires. This problem exists even for a country with a highly developed administrative infrastructure, such as the United States. Thus, a proposal by Cooper (1971 b, p. 527), for a uniform tax/subsidy scheme for goods transactions in lieu of a formal devaluation of the U. S. dollar at that time, recognized that “certain services such as air and sea transportation could also be covered by such taxes and subsidies, but administrative difficulties would dictate the exclusion of other services, e.g. personal expenditures while travelling abroad.… Administrative complexities would again dictate the exclusion of some capital transactions, notably most forms of short-term capital outflows, so these would have to be limited in some other fashion.” 15

In the four specific case studies, the turnover tax introduced by the Federal Republic of Germany posed the fewest administrative difficulties, since a system of border-tax adjustments on the destination principle had already been incorporated into the value-added tax system just installed. Thus, the import subsidy was established at 4 percentage points of the 11 per cent value-added tax for those goods subject to that rate (accounting for most goods liable for the tax), so that the net tax amounted to 7 per cent; the subsidy was 2 percentage points for goods subject to the 5.5 per cent value-added tax. Under the particular circumstances of November 1968, the authorities viewed the border-tax adjustment as the most flexible instrument available for adjusting the external imbalance. At the other extreme, the export promotion measures introduced by India, and sketched briefly in the preceding section, stand as models of administrative cumbersomeness. The industry-by-industry import entitlement schemes in particular proved unnecessarily complex and were not conducive to stimulating a sustained expansion in the production of exportable goods at competitive prices. One need not detail the various complexities of these schemes; simply note that the regulations covering the schemes for engineering goods and chemical products, for example, comprised 82 pages. Bhagwati and Desai (1970, p. 467) found that it was “difficult to escape the conclusion that, while the Third Plan witnessed a major shift towards export subsidization, export promotion policies were inefficiently designed and implemented.”

ULTIMATE DECAY OF FISCAL PROXIES

Although more difficult to document, it would seem doubtful that fiscal proxies can be implemented over long periods without causing growing corruption, evasion, or misuse, particularly when the export subsidies and import tariffs rise to high levels in lieu of a formal exchange rate change. Two particular problems from India’s experience indicate the type of dangers involved. First, Bhagwati and Desai (1970) found evidence that the import entitlement schemes led to significant overinvoicing of exports, in response to incentives created by the frequent differential between the resulting export subsidies and the black market premium on illegal foreign exchange. In their view, such over-invoicing was not necessarily harmful if it succeeded in bringing into legal channels the foreign exchange that might otherwise have remained in the illegal market, and would have been used for nonpriority purposes. On the other hand, the incentive to overinvoice apparently led some exporters, especially in such sectors as plastics, to send out shoddy goods with false, higher-price declarations, which were cleared in foreign markets at the going market rate. This practice jeopardized the buildup of good will that was regarded as essential for India’s drive to increase exports of manufactures, particularly quality and complex manufactures.

Second, an important misuse of the system grew out of the granting of rail freight concessions, largely to nonagriculture-based manufactures, to offset the transport cost “disadvantage” to exporters. As the export drive intensified, demands were made and met for freight rates that were progressively concessional as distance increased, not only for the evacuation of exports but even for the movement of raw materials to the production centers. Thus, the concessions became an important part of a policy of export maximization, in contradiction to the principle of comparative advantage. Bhagwati (1968, p. 58, footnote 27) noted that “the economic magnitude of this demand [for freight concessions] will become clear if it is realized that among those demanding such incentives from a willing Ministry were exporters of bicycles and sewing machines over 1,000 miles from the nearest outlets!”

III. Summary and Conclusions

This paper has re-examined the enduring question of whether to use the tax system as a substitute for an exchange rate change, in view of recent theoretical developments and actual country experience with fiscal proxies. The theoretical basis for a general tax/subsidy alternative to devaluation has long been established in the literature, but only in a partial-equilibrium framework. The paper has confirmed that the equivalence of the two policies remains intact in a general-equilibrium framework, as invoked by recent portfolio balance models, provided that the taxes and subsidies are extended uniformly to all foreign exchange transactions. The portfolio balance approach does, however, qualify the equivalence theorem in two important respects. First, the fiscal proxy will have a smaller impact on foreign exchange reserves—measured in domestic currency—if its financing helps to satisfy an excess demand for money (or, more generally, wealth) that would otherwise be met through the balance of payments. This would be true if the net cost of the fiscal proxy were positive and were covered by the creation of new money or assets. The distinction can, however, be overstated, since a fiscal proxy financed in this way will have an impact identical to devaluation on foreign exchange reserves measured in foreign-currency terms, which is presumably the more important measure when considering a country’s reserve position. Second, the fiscal proxy may differ from devaluation in its real effects if the tax/subsidy scheme is applied only to some subset of foreign exchange transactions, as with the traditional proposal of a uniform commercial policy for commodity trade. The creation of a dual exchange rate system in such cases means that the fiscal proxy may set in train a different adjustment process compared with devaluation; unlike devaluation, it may also have permanent real effects on resource allocation and growth, which are, a priori, uncertain.

The real issue for public policy, however, is not the theoretical possibility of equivalence between devaluation and a general tax/subsidy scheme, but the difficulty in establishing this equivalence in practice. This paper has highlighted at least four basic problems in the actual use of fiscal proxies: the tariffs and export subsidies are not comprehensive in their coverage, are often applied on a selective basis, can involve considerable administrative costs, and tend to decline in efficiency over time. The outcome in many cases has been serious distortions in the domestic price structure, maintained by an increasingly unwieldy and expensive administrative machinery—an outcome that would have been avoided by a straightforward change in the exchange rate. These difficulties were highlighted by reference to the experience of four countries that adopted fiscal alternatives of varying complexity in lieu of exchange rate action.

The important fact is that, in all four countries, the fiscal proxies were subsequently abandoned in favor of an explicit change in the exchange rate. Thus, the French system of surcharges and premiums on external transactions introduced in August 1957 was incorporated into the official exchange rate in June 1958, and this was formally devalued at the end of the year. The Federal Republic of Germany’s special turnover tax of November 1968 was temporarily suspended following the floating of the deutsche mark in September 1969, and was formally ended when the deutsche mark was revalued on October 27 of that year. In recognition of the difficulties and distortions created by its complex system of fiscal interventions, Israel undertook a substantial reform of its exchange rate system in February 1962, which involved the abolition of export premiums and import surcharges (except on luxuries), the unification of exchange rates for all merchandise trade (but not all capital transactions), and a formal 67 per cent devaluation of the Israeli pound. The dual exchange market system was retained, but operations in the securities free market were further restricted. In a similar response, India formally devalued the rupee by 36.5 per cent in June 1966; at the same time, it abolished the Tax Credit Certificate Scheme and the complex export promotion schemes, which by then covered a wide range of nontraditional exports, and undertook a substantial relaxation of restrictions on imports.

In light of this experience, it is a moot question why any government would still wish to consider a fiscal alternative to devaluation. The answer must lie in the perceived political impact of an exchange rate change. As Cooper’s (1973, p. 167) analysis of devaluation in developing countries emphasized: “Currency devaluation is one of the most dramatic—even traumatic—measures of economic policy that a government may undertake. It almost always generates cries of outrage and calls for the officials responsible to resign. For these reasons alone, governments are reluctant to devalue their currencies.” Yet, there is no reason why an equivalent fiscal proxy that was comprehensive in its coverage would not evoke the same “cries of outrage,” since its immediate impact on domestic prices is identical. And to suggest that this impact could be softened by the use of a partial fiscal proxy is to point a government back down the path toward the “patchwork efforts and marginal adjustments” 16 that characterized earlier attempts to evade exchange rate adjustment. The evidence is that such expedients may buy time in which the eventual exchange rate change may become politically acceptable or inevitable, but at a substantial efficiency cost. This cost may be minimized by a technique such as a general border-tax adjustment, as applied (although nonuniformly) by the Federal Republic of Germany to its trade in goods, but few developing countries, in particular, have the necessary administrative machinery already in place. To reemphasize, however, theoretical analysis confirms that the economic effects of devaluation can be duplicated only by a general tax/subsidy scheme applied uniformly to all foreign exchange transactions.

APPENDIX

Devaluation and Equiproportional Export Subsidies and Import Tariffs

The equivalence between devaluation and a uniform commercial policy can be demonstrated in the simple general-equilibrium model introduced by Kemp (1970) and Mundell (1971, Ch. 9), and applied to fiscal proxies by Berglas (1974) and Kuska (1976). The main assumptions of the model are as follows.

(a) The country produces one nontraded good and some traded goods, whose relative prices remain constant and are given to the country in terms of foreign currency; for this reason, traded goods can be treated as a single, composite commodity.

(b) Flexible prices and competition ensure that the economy always operates at full employment.

(c) Residents convert all foreign currency receipts and hold their wealth entirely in the form of money balances; no fractional reserve system exists, so that changes in the money stock are equal to sales or purchases of foreign currency by the monetary authority.

(d) The country has a fixed exchange rate, maintained by the monetary authority through its foreign exchange transactions.

The model thus consists of markets for a nontraded good, traded goods, and money. Under the preceding assumptions, long-run equilibrium in the economy can be summarized by the following reduced form equation system

zi=zi(p,e,M)=0i=1,2,3(1)

where z1, z2, and z3 denote excess demand for the nontraded good, traded goods, and money, respectively; p denotes the price of the nontraded good; e the exchange rate (the domestic currency price of foreign exchange); and M the quantity of money. If all prices are initially normalized at unity, e also represents the domestic price of traded goods. Given the normal homogeneity properties, the system can be rewritten

zi=zi(p/e,1,M/e)=0i=1,2,3(2)

This three-equation system has two endogenous variables—p and M. Consistency is guaranteed by the interdependence of the three equations from the aggregate budget constraint (Walras’s Law):

pz1+ez2+z3=0(3)

Consider the following two alternative policies: (a) the country devalues by 100 w per cent; (b) the country imposes a 100 w/(1 − w) uniform commercial policy that is financed by monetary means—that is, if the net outlay is positive in any period, the budget deficit is covered by newly printed money, while if the net outlay is negative, the excess money received is destroyed. With devaluation, the inverse exchange rate (1/e) decreases by 100 w per cent, equivalent to an increase in e of 100 w/(1 − w) per cent. Assume that the economy is initially in equilibrium. The increase in e leads initially to an excess demand for the nontraded good, reflecting the incentive to substitute on the demand side toward nontraded goods, and to move resources into the traded goods sector. If prices are assumed to adjust instantaneously to eliminate this excess demand, price p will rise. Given the initial stock of money balances, the higher prices for traded and nontraded goods means an excess demand for money z3, which from the aggregate budget constraint (3) is equal to the excess supply of traded goods, −ez2. That is,

ez2+z3=0(4)

Devaluation thus results initially in a balance of payments surplus B = −ez2, and the inflow of foreign exchange will begin to satisfy the excess demand for money. With perfect price flexibility, the increase in the money stock will lead to further rises in p, until long-run equilibrium is restored. In equilibrium, all prices will have increased proportionally, so that devaluation has no permanent real effects; the money stock will also have increased proportionately, in the form of an increase in foreign exchange holdings. 17

The initial impact of the uniform commercial policy is an increase in traded good prices to e' = eλ, where λ = 1 + w/(1 − w). This exogenous increase in e', by 100 w/(1 − w) per cent, is identical to that achieved by a devaluation of 100 per cent. Hence, the short-term adjustment process and the long-run equilibrium outcome is the same under both alternatives. However, the two policies have different implications for the balance of payments outturn during the adjustment process and, hence, for the net accumulation of foreign exchange. Consider the aggregate budget constraint for the economy under the uniform commercial policy, after the initial adjustment of the price for the nontraded good (for initial money balances). The constraint has the form

eλz2+z3=0(5)

The balance of payments surplus is still defined as B = −ez2, and after substitution in equation (5), this can be written as

B=1λz3(6)

This equation shows that only 1/λ of the excess demand for money during the adjustment process is satisfied by the inflow of foreign exchange; the remainder is provided by the government through the financing of the uniform commercial policy. The balance of payments (expressed in domestic currency) during the adjustment process, and the net accumulation of foreign exchange in long-run equilibrium, is therefore 100 w/(1 − w) per cent lower in absolute size under the uniform commercial policy than under devaluation. The algebraic difference in the balance of payments under the two alternatives represents the net cost of the fiscal proxy, and is given as (λ − 1)B.

In sum, the two alternative policies have exactly the same effects on the equilibrium values of all variables, as expressed in domestic currency, except (a) the initial level of foreign exchange holdings, (b) the balance of payments during the adjustment process, and (c) the net accumulation of foreign exchange reserves. However, when expressed in foreign currency, both the balance of payments and the initial and final foreign exchange position are invariant under the two policies.

BIBLIOGRAPHY

  • Andel, Norbert, Zum Gesetz über massnahmen zur aussenwirtschaftlichen Absicherung,” Finanzarchiv, Vol. 28 (August 1969), pp. 40114.

    • Search Google Scholar
    • Export Citation
  • Baruch, Joseph, Import Taxes and Export Subsidies in Israel, 1955–61,” Bank of Israel Bulletin, No. 18 (March 1963), pp. 4870.

    • Search Google Scholar
    • Export Citation
  • Berglas, Eitan, Devaluation, Monetary Policy, and Border Tax Adjustments,” Canadian Journal of Economics, Vol. 7 (February 1974), pp. 111.

    • Search Google Scholar
    • Export Citation
  • Bhagwati, Jagdish N., The Theory and Practice of Commercial Policy: Departures from Unified Exchange Rates, Special Papers in International Economics, No. 8, International Finance Section, Princeton University (January 1968).

    • Search Google Scholar
    • Export Citation
  • Bhagwati, Jagdish N, and Padma Desai, India: Planning for Industrialization—Industrialization and Trade Policies Since 1951, Organization for Economic Cooperation and Development, Development Centre (Oxford University Press, 1970).

    • Search Google Scholar
    • Export Citation
  • Boyer, Russell S. (1977 a), “Devaluation and Portfolio Balance,” American Economic Review, Vol. 67 (March 1977), pp. 5463.

  • Boyer, Russell S. (1977 b), “Commercial Policy under Alternative Exchange Rate Regimes,” Canadian Journal of Economics, Vol. 10 (May 1977), pp. 21832.

    • Search Google Scholar
    • Export Citation
  • Cooper, Richard N. (1971 a), “Devaluation and Aggregate Demand in Aid-Receiving Countries,” in Trade, Balance of Payments, and Growth: Papers in International Economics in Honor of Charles P. Kindleberger, ed. by Jagdish N. Bhagwati, Ronald W. Jones, Robert A. Mundell, and Jaroslav Vanek (Amsterdam, 1971), pp. 35576.

    • Search Google Scholar
    • Export Citation
  • Cooper, Richard N. (1971 b), “The Nexes among Foreign Trade, Investment, and Balance- of-Payments Adjustment,” in United States International Economic Policy in an Interdependent World, papers submitted to the U. S. Commission on International Trade and Investment Policy, Compendium of Papers: Vol. 2 (Washington, July 1971), pp. 51538.

    • Search Google Scholar
    • Export Citation
  • Cooper, Richard N. (1973), “An Analysis of Currency Devaluation in Developing Countries,” Ch. 9 in International Trade and Money: The Geneva Essays, ed. by Michael B. Connolly and Alexander K. Swoboda (London, 1973), pp. 16796.

    • Search Google Scholar
    • Export Citation
  • Dornbusch, Rudiger (1974), “Real and Monetary Aspects of the Effects of Exchange Rate Changes,” in National Monetary Policies and the International Financial System, ed. by Robert Z. Aliber (University of Chicago Press, 1974), pp. 6481.

    • Search Google Scholar
    • Export Citation
  • Dornbusch, Rudiger (1975 a), “A Portfolio Balance Model of the Open Economy,” Journal of Monetary Economics, Vol. 1 (January 1975), pp. 320.

    • Search Google Scholar
    • Export Citation
  • Dornbusch, Rudiger (1975 b), “Exchange Rates and Fiscal Policy in a Popular Model of International Trade,” American Economic Review, Vol. 65 (December 1975), pp. 85971.

    • Search Google Scholar
    • Export Citation
  • Fleming, J. Marcus, Dual Exchange Markets and Other Remedies for Disruptive Capital Flows,” Staff Papers, Vol. 21 (March 1974), pp. 127.

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., and Harry G. Johnson (eds.), The Monetary Approach to the Balance of Payments (London, 1976).

  • Frenkel, Jacob A., and Carlos A. Rodriguez, Portfolio Equilibrium and the Balance of Payments: A Monetary Approach,” American Economic Review, Vol. 65 (September 1975), pp. 67488.

    • Search Google Scholar
    • Export Citation
  • Goode, Richard, Comparative Economic Effects of Multiple Exchange Rates, Export Taxes, Production Taxes, and Profits Taxes” (unpublished, International Monetary Fund, April 30, 1952).

    • Search Google Scholar
    • Export Citation
  • Great Britain, Treasury, Committee on Finance and Industry, Report, Cmd. 3897 (London, 1931), pp. 199200.

  • Haberler, Gottfried, Taxes on Imports and Subsidies on Exports as a Tool of Adjustment,” in Monetary Problems of the International Economy, ed. by Robert A. Mundell and Alexander K. Swoboda (University of Chicago Press, 1969), pp. 17379.

    • Search Google Scholar
    • Export Citation
  • Hahn, Frank H., The Monetary Approach to the Balance of Payments,” Journal of International Economics, Vol. 7 (August 1977), pp. 23149.

    • Search Google Scholar
    • Export Citation
  • Hicks, J.R., Free Trade and Modern Economics,” reprinted as Ch. 3 in his Essays in World Economics (Oxford University Press, 1959), pp. 4065.

    • Search Google Scholar
    • Export Citation
  • Isard, Peter, Exchange-Rate Determination: A Survey of Popular Views and Recent Models, Princeton Studies in International Finance, No. 42, International Finance Section, Princeton University (1978).

    • Search Google Scholar
    • Export Citation
  • Johnson, Harry G. (1972), “The Monetary Approach to Balance-of-Payments Theory,” Ch. 9 in his Further Essays in Monetary Economics (London, 1972), pp. 22949.

    • Search Google Scholar
    • Export Citation
  • Johnson, Harry G. (1977), “The Monetary Approach to Balance of Payments Theory and Policy: Explanations and Policy Implications,” Económica, Vol. 44 (August 1977), pp. 21729.

    • Search Google Scholar
    • Export Citation
  • Jones, Ronald W., Trade with Non-traded Goods: The Anatomy of Interconnected Markets,” Económica, Vol. 41 (May 1974), pp. 12138.

    • Search Google Scholar
    • Export Citation
  • Kemp, Murray C, The Balance of Payments and the Terms of Trade in Relation to Financial Controls,” Review of Economie Studies, Vol. 37 (January 1970), pp. 2531.

    • Search Google Scholar
    • Export Citation
  • Kuska, Edward A., Devaluation, Equi-proportional Export Subsidies and Import Tariffs, and Transfer Payments,” Económica, Vol. 43 (May 1976), pp. 18184.

    • Search Google Scholar
    • Export Citation
  • Lanyi, Anthony, Separate Exchange Markets for Capital and Current Transactions,” Staff Papers, Vol. 22 (March 1975), pp. 71449.

  • McKinnon, Ronald I. (1969), “Portfolio Balance and International Payments Adjustment,” in Monetary Problems of the International Economy, ed. by Robert A. Mundell and Alexander K. Swoboda (University of Chicago Press, 1969), pp. 199234.

    • Search Google Scholar
    • Export Citation
  • McKinnon, Ronald I. (1973), Money and Capital in Economic Development, The Brookings Institution (Washington, 1973), Ch. 10.

  • Meade, J.E., The Balance of Payments: The Theory of International Economic Policy, Vol. 1 (Oxford University Press, 1951).

  • Mundell, Robert A. (1961), “Flexible Exchange Rates and Employment Policy,” Canadian Journal of Economics and Political Science, Vol. 27 (November 1961), pp. 50917.

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A. (1971), Monetary Theory: Inflation, Interest, and Growth in the World Economy (Pacific Palisades, California, 1971).

  • Whitman, Marina V.N., Global Monetarism and the Monetary Approach to the Balance of Payments,” Brookings Papers on Economic Activity: 3 (1975), pp. 491536.

    • Search Google Scholar
    • Export Citation
*

Mr. Laker, economist in the African Department, was in the Fiscal Affairs Department when this paper was written. He is a graduate of the University of Sydney and the London School of Economics and Political Science. Before joining the Fund, he was an economist at the Australian Treasury.

1

“This proposal would avoid the injury to the national credit and to our receipts from foreign loans fixed in terms of sterling which would ensue on devaluation.”

2

Meade (1951, pp. 268–72); Goode (1952).

3

See Johnson (1972) and Isard (1978) for a fuller description of the “standard” or “popular” balance of payments view.

4

See Cooper (1971 a) for a general treatment of the elasticity conditions. A formal analysis of the general equilibrium implications of the elasticities approach is given in Dornbusch (1975 a).

5

This is consistent with Mundell’s (1961, p. 514) definition of commercial policy as “any policy which restricts imports or promotes exports without directly affecting the level of saving or investment.”

6

Throughout the paper, the term “portfolio balance approach” refers to those models that focus on the requirement that stocks of money and other financial assets equal stock demands for these assets as a necessary condition for equilibrium. This approach is a generalization of the so-called monetary approach to the balance of payments, which concentrates on only one asset market—that for money balances. For explanations of the monetary approach, see Johnson (1972; 1977) and Whitman (1975).

7

There may also be a difference—in the demand for money—between the two alternatives. Under a uniform commercial policy, the government is collecting and then redistributing the tax proceeds and subsidy payments; this action could increase the transactions demand for money.

9

See Hahn (1977), especially.

10

Hahn (1977, p. 245), for example, has argued that “of all the forces to be considered [in devaluation] the real cash balance effects seem the least, not the most, important.” See Whitman (1975, p. 525) for references to empirical work on this question. Frenkel and Johnson (1976) also contains a number of empirical studies.

13

A description of this scheme, which highlights the departure from the uniformity principle, is given in Andel (1969), especially p. 407.

14

This system and its effects are discussed at length in Bhagwati and Desai (1970).

15

In fact, a form of fiscal proxy for devaluation was an integral part of the New Economic Policy introduced by the United States on August 15, 1971. Imports were subject to a temporary surcharge, generally at the rate of 10 per cent, although the surcharge extended to only about one half of U. S. imports. Incentives for exports took the form of tax deferments for new corporations, known as Domestic International Sales Corporations, to ensure that U. S. exporters from domestic sources would be able to enjoy tax treatment comparable to that of U. S. producers located abroad. See Economic Report of the President (Washington, January 1972).

17

In the more realistic case, devaluation will occur from a position of balance of payments deficit, involving an initial excess supply of money and excess demand for traded goods. Devaluation will mop up part, all, or more than all of the initial excess supply of money, so that the net increase (positive or negative) in the quantity of money demanded to restore real balances will be less than proportional to the devaluation.