This paper examines the distinctive features of the exchange rate systems of planned economies and the role that exchange rates might play in economic stabilization and structural adjustment in these countries. That this issue has been relatively neglected is understandable in light of the conventional wisdom that the exchange rate in the classical centrally planned economy (CPE) has little or no effect on prices or real trade flows, and has therefore essentially only an accounting function (see Pryor (1963), Holzman (1968) and Wolf (1980a)).

It is true that in recent years, mainly in response to the major decentralization of foreign trade in Hungary, some attention has been devoted to the adjustment role of the exchange rate in “modified” planned economies (MPEs). However, most analyses of this issue have been either overstylized, or so country-specific as virtually to preclude generalization (see, respectively, Wolf (1978); and Portes (1979), Kozma (1981), and Marer (1981a, 1981b)). Yet another strand of literature, especially prevalent in the 1960s and early 1970s, surveyed the evolution of an almost bewildering array of “effectiveness indices” for foreign trade that, in the absence of economically meaningful official exchange rates, were being used with varying degrees of success to improve the commodity and geographical structure of CPE foreign trade.¹ The main focus of these studies, however, was the planning of foreign trade rather than the implications of such planning devices for shorter-run stabilization or longer-run structural adjustment of the economy as a whole.

In Sections I–III of this paper, a common framework is developed to facilitate the comparative analysis of exchange rates between stylized market and planned economies and among the planned economies. This framework is used in Sections IV-VII to examine the exchange rate regimes, and to a certain extent, policies of four members of the Council for Mutual Economic Assistance (CMEA): the U.S.S.R., which has a relatively classical CPE system; the German Democratic Republic; Poland (in the 1970s only); and the MPE of Hungary. In these sections, there is an attempt to avoid the opposite pitfalls of overstylization and excessive institutional detail. The diverse exchange rate concepts, systems, and practices among these four countries, and their recent evolution, are discussed, and emphasis is laid on many of the common problems their policymakers face in attempting to enhance the role of exchange rates. Also explored are both the theoretical and practical links between exchange rates and variables controlled by foreign trade decision makers, and the extent to which the latter have the incentive and the authority to respond to changes in exchange rates. The exchange rate systems of these countries probably encapsulate the range of systems now existing within the CMEA, and this comparative analysis, without being comprehensive, attempts to illuminate the actual and potential functions that exchange rates may perform in planned economies.²

I. The Market Economy

The analysis of the role of the exchange rate in stylized models of market economies usually begins by assuming that most or all foreign trade is carried out by autonomous profit-maximizing firms. Decisions to export or import are made on the basis of profitability calculations in terms of domestic currency costs and prices. International commodity arbitrage is assumed, and as a result, the following familiar equilibrium condition holds for the ith product (ignoring transport costs and trade commissions):

where P_i and $P_i^{*}$ denote, respectively, the domestic currency price and the foreign currency price of the ith traded good, e is the official exchange rate, expressed as the domestic currency price of foreign exchange, and t_i denotes the ad valorem trade tax (or subsidy) levied by the government on the i th good.

Equation (1) is an equilibrium condition because any differences between the actual domestic currency price of the ith product and the domestic currency equivalent of the foreign price (plus the trade tax) will present the private sector with arbitrage profit opportunities, and the arbitrage itself will eliminate the price differential.

In many market economies, explicit trade taxes are often replaced or supplemented by quantitative restrictions on trade, payments, or both. A binding quantitative restriction on the ith product will drive a further wedge between the two prices in equation (1). Although in this case the government is (via these controls) directly influencing the volume of foreign trade, and indirectly influencing the domestic price, the actual levels of domestic production and consumption, as well as the domestic price, will still be determined by the market. Even if the government intervenes with controls on trade, payments, or both, the exchange rate in a market economy still provides critical signals for domestic resource allocation. And even if the exchange rate is freely determined in the market for foreign exchange, its level and therefore calculations by the private sector of the gains from trade will be significantly influenced by the pattern of explicit trade taxes and subsidies and direct controls on trade.

One important role usually ascribed to the exchange rate in market economies is as a transmitter of price signals from the world market, reflecting shifts in demand patterns and the costs of production. The greater the role permitted for international commodity arbitrage in the domestic economy, the more efficiently will the exchange rate be able to perform this role in transmitting information.

Changes in the exchange rate are commonly thought to affect short- and medium-term resource allocation and macroeconomic balance in the stylized market economy. This occurs because of the interaction of aggregate expenditure-reducing effects and expenditure-switching responses to changes in relative prices. By raising the domestic prices of tradable goods, devaluation reduces both the real wage and real money balances, unless this effect is immediately offset by proportionate increases in the nominal wage and the nominal supply of money. The decline in real wages and real balances will lead to a fall in aggregate real expenditure, and this will keep the price of nontradable goods from rising proportionately to the price of tradables, at least in the short and medium run. This increase in the relative price of tradables will encourage a reallocation of resources into the production of tradables and the consumption of nontradables, with attendant positive effects on the balance of trade. If the country is “large” with respect to at least one of its tradables, devaluation will also in general affect the terms of trade, which will further influence both the balance of trade and the structure of domestic production.

Krueger (1978) has also emphasized the impact that devaluation may have on the relative prices of tradables, even in a “small” country, if devaluation is accompanied by the relaxation or elimination of binding, quantitative restrictions on imports. Whereas the domestic price of the exportable goods will increase proportionately to the exchange rate, relaxation of restrictions will permit the domestic price of the importable goods to fall. The degree of bias toward import substitution will be reduced and productive resources will be reallocated into exportables.

II. The Centrally Planned Economy

Whereas efficiency considerations are paramount in private decisions regarding foreign trade in the market economies, they have played a decidedly secondary role in the classical CPE. In the CPE, all activities of the state-owned (or socialized) sector are subject to detailed central planning. Foreign trade is carried out by noncompeting state-owned foreign trade organizations that are organizationally separate from domestic industrial enterprises and wholesalers, and are subordinate to the Ministry of Foreign Trade. The foreign trade plan is an integral part of the overall national economic plan. The plan for imports essentially reflects the pattern of planners’ excess demands for domestically produced goods in the priority sectors. Exports are planned in light of the foreign exchange needs implied by the import plan and a balance of payments constraint.³

Although the foreign trade organizations may have some material incentive to obtain high foreign currency prices for exports and to pay low foreign currency prices for imports, there is little evidence that they themselves actively manipulate trade volumes in response to or to affect foreign currency prices. There are several reasons for this, including the incorporation of foreign trade volumes into the complex balancing of quantities in the national economic plan, and bilateral trade agreements with other planned economies that specify the volume as well as the price of deliveries on an annual and multiyear basis. Historically, the rather inward-looking orientation of foreign trade decision makers in CPEs, combined with the strict institutional segregation of production and foreign trade activities, have also made the trade organizations relatively insensitive to short-run movements in world market prices.⁴

Given that foreign trade in the classical CPE is driven by the planners’ demand for imports, the planners themselves might be expected to be sensitive to changes in the overall terms of trade. As discussed in Holzman (1968), a deterioration in the terms of trade, combined with a trade balance constraint, might lead the planners to expand rather than contract exports, as would be expected of profit-oriented market economy firms. Wolf (1982a) formalized this argument, and also found empirical evidence for such a backward-bending short-run offer curve for Soviet trade with member countries of the Organization for Economic Cooperation and Development (OECD) in the 1970s. Significant short-run deviations in the plan for export volumes are therefore probably only decided upon at the higher levels in the planning hierarchy, in response to trends in the overall terms of trade and balance of trade.

Another distinctive feature of the classical CPE is the attempt to insulate domestic prices from price movements on the world market. Domestic prices within the socialized sector of the CPE are administratively fixed, and are usually held constant for long periods. The official exchange rate applied to commercial transactions is typically established by the authorities with reference to historical gold parities, or on some other basis little related to the CPE’s foreign trade position. As a group, the foreign trade organizations earn a profit (F) from foreign trade equal to:

where B_T is the trade balance evaluated in domestic prices and $B_T^{\prime }$ is the so-called valuta (or deviza) trade balance. The latter is defined as the trade balance evaluated in foreign currency prices multiplied by the official exchange rate applied to commercial transactions, or e΄. Because the resultant valuta prices ($P_i^{\prime }=P_i^{*}{e}^{\prime }$ for the ith traded good) have no connection with the prices prevailing on domestic markets, this exchange rate (e΄) is sometimes referred to as the official external (or valuta) exchange rate. (See Table 1 for a glossary of important prices and exchange rates in planned economies.)

Table 1.

Glossary of Important Prices and Exchange Rates in Planned Economies

Table 1.

Glossary of Important Prices and Exchange Rates in Planned Economies

(1 + αi)	=	Implicit internal exchange rate for the ith good ($P_i/P_i^{\prime }$) see equation (4).
e	=	Exchange rate in the market economy; see equation (1).
e’	=	Official external (or valuta) exchange rate (for instance, units of valuta rubles per dollar); see equation (3).
e"	=	Official internal exchange rate (for example, units of domestic or internal rubles per valuta ruble); see equation (4).
$e_s^{"}$	=	Shadow (“optimal”) internal exchange rate; see Figure 1.
$e_{xa}^{"},e_{ma}^{"}$	=	Average real implicit internal exchange rate for exports (average Zxi/$P_{xi}^{\prime }$) and imports (average Zmi/$P_{mi}^{\prime }$) respectively; see Figure 1.
exa=		Average domestic cost of earning one unit of foreign exchange (average Pxi/$P_{xi}^{*}$).
Pi	=	Domestic price at which the ith good is transferred between foreign trade organizations and domestic industrial or wholesale enterprises; see equations (3) and (4). (Also denotes the domestic price of the ith good in the market economy; see equation (1).)
$P_{xi}^{*}$	=	Foreign currency price of the ith good.
$P_i^{\prime }$	=	Valuta price of the ith good ($P_i^{*}{e}^{\prime }$); see equation (3).
$P_i^{"}$	=	Calculative or transactional price of the ith good; see equation (4).
ti	=	Explicit ad valorem trade tax or subsidy rate on the ith good; see equations (1) and (4).
Zi	=	Estimated “economic value,” or shadow price, of the ith good, in domestic currency; see Figure 1.

View Table

Table 1.

Glossary of Important Prices and Exchange Rates in Planned Economies

(1 + αi)	=	Implicit internal exchange rate for the ith good ($P_i/P_i^{\prime }$) see equation (4).
e	=	Exchange rate in the market economy; see equation (1).
e’	=	Official external (or valuta) exchange rate (for instance, units of valuta rubles per dollar); see equation (3).
e"	=	Official internal exchange rate (for example, units of domestic or internal rubles per valuta ruble); see equation (4).
$e_s^{"}$	=	Shadow (“optimal”) internal exchange rate; see Figure 1.
$e_{xa}^{"},e_{ma}^{"}$	=	Average real implicit internal exchange rate for exports (average Zxi/$P_{xi}^{\prime }$) and imports (average Zmi/$P_{mi}^{\prime }$) respectively; see Figure 1.
exa=		Average domestic cost of earning one unit of foreign exchange (average Pxi/$P_{xi}^{*}$).
Pi	=	Domestic price at which the ith good is transferred between foreign trade organizations and domestic industrial or wholesale enterprises; see equations (3) and (4). (Also denotes the domestic price of the ith good in the market economy; see equation (1).)
$P_{xi}^{*}$	=	Foreign currency price of the ith good.
$P_i^{\prime }$	=	Valuta price of the ith good ($P_i^{*}{e}^{\prime }$); see equation (3).
$P_i^{"}$	=	Calculative or transactional price of the ith good; see equation (4).
ti	=	Explicit ad valorem trade tax or subsidy rate on the ith good; see equations (1) and (4).
Zi	=	Estimated “economic value,” or shadow price, of the ith good, in domestic currency; see Figure 1.

The foreign trade profit of the foreign trade organizations is essentially nominal from their standpoint, because the government automatically taxes away profits and subsidizes losses. In effect, this variable tax-subsidy process, called price equalization, severs the links between foreign currency and domestic wholesale prices. Because of the price equalization process, Fin equation (2) is analogous to the net trade tax revenue collected by the government in a market economy (see Wolf (1980b)).

As long as domestic prices and foreign trade volumes (and therefore necessarily foreign currency prices of traded goods) are assumed to be invariant with respect to changes in the external exchange rate, its modification by the authorities will at most only affect the nominal profits of the trade organizations. The stylized CPE model therefore yields the conclusion that the exchange rate has only an accounting role in such economies (Wolf (1980a)). (Various analysts, such as Holzman (1968) and van Brabant (forthcoming), have suggested that foreign trade organizations may not be totally indifferent to the level of the official external exchange rate. But even they have suggested that this exchange rate plays at most only a negligible role in classical CPEs.)

The fundamental insulation of domestic prices of traded goods from foreign currency prices may be summarized by:

where $P_i^{*}$ and e΄ are defined as before, P_i is the domestic currency price at which the ith product is transferred between foreign trade organizations and domestic enterprises, and α_i represents the ex post commodity-specific variable implicit tax rate on the ith product. With the domestic price and the external exchange rate fixed for long periods by the authorities, α_i will vary with fluctuations in the foreign currency price of the ith good that is paid or received by the trade organizations. The variable (1 + α_i) is simply the ratio of the domestic price to the so-called valuta price of the ith product, and may be thought of as the commodity-specific implicit internal exchange rate.⁵

The familiar and relatively straightforward relationships between prices and opportunity costs in the stylized market economy tend to disappear in the CPE. The opportunity cost evaluations implicit in the domestic supply and demand curves underlying foreign exchange market analysis for the market economy are not in general reflected in the administratively set prices of the CPE. Domestic producer prices are typically based on the historical average cost of production (sebestoimosf) of the relevant industrial branch, and are held relatively fixed over time and over fairly wide ranges of output.

Aside from the lack of a good measure of opportunity cost on the domestic price side, the planners soon recognize that, unlike that of the market economy, their official external exchange rate embodies no economically useful information. Only by coincidence would this rate even approximate the domestic accounting cost of earning foreign exchange, let alone the real cost. Indeed, the external exchange rates of most CPEs have historically understated the domestic currency cost of earning foreign exchange. Fairly early on, therefore, many planners realized that they needed both to improve their measures of domestic costs and use values of products, and to adjust the official external exchange rates to reflect the true cost of foreign exchange more accurately. Clearly, the success of the second endeavor depended very much on progress with respect to the first.⁶

One approach taken in the late 1950s and the 1960s to improve the efficiency of foreign trade planning in the CPEs involved the development of indices of “foreign trade effectiveness” (Boltho (1971), and Hewett (1974)). One type of index, established by the authorities in several CMEA countries, was the so-called official internal exchange rate; it was designed to compensate for the arbitrariness of the external exchange rate and to influence the pattern and possibly even the volume of trade. An internal exchange rate—or units of internal domestic currency per unit of external or valuta domestic currency—could be established for trade as a whole (e"), or internal rates could be differentiated by industrial branch or commodity groups. The official internal exchange rate roughly bridges the gap between the average domestic price of traded goods in a particular group and their average valuta price. As a benchmark ratio of domestic-to-valuta prices, the official internal rate is the basis for the calculative price ($P_i^{"}$) of the ith product:

Here all the variables on the right-hand side are as defined earlier. The calculative price for the ith product would only coin-cidentally be equal to the prevailing domestic wholesale price (Pi). The calculative price becomes a standard against which the domestic wholesale price of a product can be compared. If the calculative price is greater than the prevailing domestic price, for example, this might suggest to the foreign trade planner that this product could be a profitable export. Because the domestic price would only coincidentally reflect the real marginal cost of domestic production, however, the usefulness of such official internal exchange rates is limited.

Recognition of these limitations led to early attempts, especially by Hungarian and Polish economists, to develop shadow exchange rate measures using large-scale linear programming models.⁷ The basic idea is to optimize some objective function (which could be to minimize prime cost or maximize consumption), subject to linear constraints with respect to production capacities and the balance of trade. Production functions with fixed coefficients, and constant costs, both for domestic- and foreign-produced goods, are assumed. The optimization process yields requirements for real trade flow requirements by commodity, and shadow prices in domestic currency for each tradable good.

In Figure 1, importables are arranged in descending order by the calculated ratio of their domestic shadow price (narodnokho-ziaistvennaia otsenka, or Z_mi) to the valuta equivalent of their foreign currency price ($P_{mi}^{\prime }$). Exportables are arranged in ascending order of the ratio of their real domestic cost of production (Z_xi) to their valuta price ($P_{xi}^{\prime }$). (If “full” shadow exchange rates are sought, the (Z_i/$P_i^{*}$) are plotted against $V_i^{\prime }=P_i^{*}{e}^{\prime }{Q}_i,$.) The horizontal axis measures the cumulative value of these imports and exports in valuta; in other words, $V_i^{\prime }=e^{\prime }{Q}_i,$, where Q_i denotes the volume of the ith good traded. Because constant marginal costs are assumed for each product, the valuta expenditure ($V_m^{\prime }$) and revenue ($V_x^{\prime }$) curves in the figure are really just approximations to bar graphs composed of individual commodities (such as importables a, b and c, and exportables d and f).

View Full Size

The “Optimal” Internal Exchange Rate in a Planned Economy

Citation: IMF Staff Papers 1985, 002; 10.5089/9781451946932.022.A002

• Download Figure
• Download figure as PowerPoint slide

According to this approach, imports are only profitable if the real saving of resources they entail or their use value to the national economy is greater than the real value of the resources used to earn the foreign exchange for their purchase. This suggests that trade should be expanded, in the order indicated in the diagram, only as long as ($\left(Z_{mi}/P_{mi}^{\prime }\right)>(Z_{xi}/P_{xi}^{\prime })$)—so, for example, a, b and c should be imported and d and f exported. Thus the “optimal” level of balanced trade is $V_0^{\prime }$ at the intersection of the $V_m^{\prime }$ and $V_x^{\prime }$ curves. This intersection also yields the shadow internal exchange rate $e_s^{"}$, at which the domestic (average) real relative price of the marginal tradables is equal to their relative price in the foreign market. The net gain to the economy, when trade is conducted at $V_0^{\prime }$, can be measured by the triangle ABC. It can easily be shown that a different pattern of trade, whether balanced or not, would yield a smaller sum of differential rents. (For more detail, see Trzeciakowski (1978) and Shagalov (1973).)

The figure must not be confused with the foreign exchange market diagram typically constructed for the market economy. (The same proviso applies to interpretation of the counterpart of Figure 1 plotted for Zi/$P_i^{*}$.) The demand schedule for foreign exchange in the market economy, for instance, indicates the aggregate demand of decentralized economic agents for foreign exchange, at various exchange rates that are, from their standpoint, exogenously determined. In the figure, the vertical axis measures Zi/$P_i^{\prime }$, or what can be considered as the real implicit internal exchange rate of the ith product in a CPE. The $V_m^{\prime }$ schedule in effect reflects the planners’ demand for valuta at different shadow internal exchange rates. It is the planners themselves, however, who will determine the actual shadow internal exchange rate. Moreover, it is their own preferences, implicitly reflected in the V’ schedules, that will influence their ultimate choice of this exchange rate.

Another important distinction is that the foreign exchange market schedule for the market economy is the summation of interdependent foreign exchange curves relating to individual commodities. Compare, for example, the impact on real imports of increasing the market economy’s exchange rate, and raising the official internal rate for the CPE in the figure from $e_0^{"}$ to, say, $e_1^{"}$. In the latter case, the planners would want to eliminate entirely imports of product c, but items b and a, with high calculated real implicit internal rates, would continue to be imported in full measure. In the market economy, on the other hand, generally real imports of all products would be contracted by an amount that varied according to their different price elasticities of import demand. This basic difference reflects, of course, the underlying assumption of constant costs and use values in the optimization approaches of the East European and Soviet economists.

Observe also that the shadow internal exchange rate in the figure is not really an “equilibrium” rate in its usual, market sense. Were the foreign trade planners to relinquish direct control over allocation of resources, trade, and foreign exchange, there would be no reason to expect that market forces would lead to an equilibrium internal exchange rate at $e_s^{"}$. Although the shadow rate would lead to balanced trade in a CPE (and reflects an “equilibrium” level of trade in that sense), trade could also be balanced at levels other than $V_0^{\prime }$ in the figure, by using different internal exchange rates for imports and exports. Or if the target were, say, a trade deficit, the shadow rate would be below the level shown in the figure. The shadow rate in the figure is an “optimal” rather than an “equilibrium” exchange rate, given domestic shadow prices and valuta prices.

III. The Modified Planned Economy

The principal distinguishing features of the MPE are set forth in Wolf (1985). The most significant features, from the standpoint of the exchange rate system, are the elimination of detailed central planning of enterprise inputs, outputs, and foreign trade; the designation of enterprise profitability rather than plan fulfillment as the major incentive criterion; a significantly increased scope for price-setting by many enterprises, on more or less free markets; and the encouragement of more direct, export-oriented links between domestic enterprises and foreign markets, as well as the establishment of links between foreign currency and domestic prices for many products. Many of these characteristics of MPEs, as well as the continuing influence of inherited policy and systemic features of the CPE, are discussed in greater detail in Section VII on Hungary.

In the MPE the calculative price ($P_i^{"}$) in equation (4) becomes a transactional price. This means, for example, that the exporting enterprise, whether a foreign trade organization or an industrial enterprise with direct foreign trade rights, will now receive $P_i^{"}$ units of domestic currency for each unit of the ith product that is sold abroad. Unlike the foreign trade organizations in the classical CPE, for which valuta earnings have no domestic purchasing power, the exporting enterprise in the MPE actually converts its foreign exchange earnings (at the official “full” exchange rate, e’e") into a domestic currency that can be used on domestic markets. Moreover, as discussed in some detail in the section on Hungary, this transaction price is in many cases designed to affect, either directly or indirectly, the price at which the traded good actually changes hands domestically.

In the MPE the full exchange rate is roughly analogous to the rate for the stylized market economy given in equation (1). This full exchange rate is variously referred to as a foreign trade coefficient or multiplier, as it was in Hungary between 1968 and 1975, or as the commercial exchange rate, as in Hungary and Poland beginning in 1976 and 1982, respectively.

IV. The Union of Soviet Socialist Republics

The organization of decision making for foreign trade in the Soviet Union is documented in considerable breadth and detail in Hewett (1974), Gruzinov (1979), and Gardner (1983). The actual degree of autonomy possessed by foreign trade organizations in the U.S.S.R. remains, however, unclear, as does the pattern of formal authority. The reforms of the Soviet foreign trade system of 1978–79 aimed to improve coordination between the Ministry of Foreign Trade, to which virtually all foreign trade organizations are subordinate, and the relevant industrial ministries. New material incentive funds were also established for the trade organizations, and various internal changes were made in their structure to promote coordination with the industrial enterprises. This partial reorganization of the foreign trade system may have improved the coordination and effectiveness of foreign trade. Nevertheless, it is not clear whether the indicators used by the Central Planning Committee (Gosplan) of the economic effectiveness of foreign trade (see below) are as yet considered to be obligatory indicators either for the trade organizations or for their domestic customers or suppliers (VAVT (1983)).

The planners have been most concerned with two measures of foreign trade effectiveness. One measure is budgetary effectiveness, and this is summarized by the nominal “profit” earned by the trade organizations (see equation (2)). Budgetary effectiveness is a good approximate measure of the CPE’s static gain from foreign trade if the domestic cost structure reflects relative scarcities and valuta trade is balanced. But this measure provides no information about whether the level or composition of trade is at or near the optimum.

It is quite possible that, in contrast to the stylized view of the classical CPE, the Soviet trade organizations themselves have an interest in the size of these profits. The commissions and payments made into their bonus funds are related to their valuta turnover, but they are also rewarded on the basis of above-plan convertible-currency export earnings and their economizing on hard-currency expenditures (see Zakharov and Shagalov (1982) and VAVT (1983), pp. 88-89). If the trade organizations are interested in profits, they will also be interested in the official external exchange rate. If they have no discretion on trade volumes, whether their profits increase with a rise in the external exchange rate will depend on whether their initial valuta balance is positive. If, however, the trade organizations do have some discretion with respect to trade volumes, then an interest in maximizing, or at least increasing, “profitability” would lead them to increase real exports in the event of an increase in the external exchange rate. Rather than seek to reduce real imports, however, which now would be less “profitable” after devaluation, they might attempt to increase them so as to offset the decline in the difference between the fixed domestic wholesale price and the valuta price. This in itself might be sufficient reason for the authorities to limit severely the discretion of the foreign trade organizations with respect to import trade volumes.⁸

Although the trade organizations are probably severely limited in their ability to manipulate real trade flows over the period of a given plan, they probably do have some say (because of their knowledge of market conditions) in the initial development of the foreign trade plan. If they care about their foreign trade profits, they would have an interest in shifting exports towards regions for which the authorities had established relatively high external exchange rates, and emphasizing imports from areas for which rates were low.

Whatever logic holds for the response to external exchange rates should also hold for movements in foreign currency prices. If the trade organizations are able to influence real trade flows, at least in the planning process, they should be expected to push for increased real exports as valuta export prices rise. Yet Wolf (1982a) has shown that Soviet real exports to the West between 1970 and 1978 were actually reduced, ceteris paribus, as Soviet terms of trade improved due to significant increases in export prices.⁹ This suggests that either the trade organizations had little material interest in their nominal foreign trade profits, or that if they did, they had little influence over the real trade flows that partly determined these profits.

The case for moving beyond calculations of budgetary effectiveness to a measure of valuta or economic effectiveness has been discussed. The Soviet economists Shagalov and Faermark (1981) argue for the use of an “optimal” or “limit” internal exchange rate and observe (on p. 11) that:¹⁰ “The necessity for utilizing the above-mentioned calculated indicators is due to the fact that the operative exchange rate, reflecting the gold content of the ruble, does not give an adequate valuation to the economic significance of this or that currency, and prices, operating in domestic trade, in many cases inadequately reflect the socially required expenditures involved in their production.”

Official internal exchange rates have been used in Soviet foreign trade planning for years. Different rates for various groups of imported machinery and equipment have been used, for example, since the 1950s (Zakharov (1982)). These rates were fixed for a number of years on the basis of the average implicit internal exchange rate in the base year for a given group of products. The resulting calculative prices (see equation (4)), were not only designed to influence decision making regarding whether to import, but are also said to have directly or indirectly influenced overall producer prices. It is not at all clear, however, whether this particular price linkage for imports of machinery and equipment, provided by the official internal exchange rate, is more than an anomaly in an otherwise relatively insulated economy.¹¹

Debate continues in the Soviet Union about the most appropriate measure of the economic effectiveness of foreign trade. One controversy is over the definition of the domestic prices that should enter into the calculations.¹² At issue is whether existing wholesale prices, which have historically been closely related to average costs, so-called full cost (privedennye zatrat), which includes a capital charge, or more sophisticated shadow prices should be used.

Mathematical economists in the Soviet Union have refined the methodology for calculating shadow exchange rates (along the lines of the figure), and have also presented evidence suggesting the lack of systematic correlation among wholesale prices, calculated full costs, and “economic values” (or shadow prices—see Shagalov (1983)). As might be expected in an economy without market-clearing prices in general, the definition of “economic value” of tradables in the U.S.S.R. depends on whether they are considered “deficit” or “nondeficit” goods. Zakharov and Shagalov (1982) consider the economic value of the nondeficit goods to be simply the full cost involved in their production—or average branch costs plus a capital charge. The export of deficit goods requires domestic consumers to substitute other products that are presumably higher cost or less efficient, or both. For deficit exportables, the economic value is defined as the full cost of the domestic substitute plus the additional costs borne by consumers in having to make the substitution. On the import side, economic value is defined as the full cost of the domestic import substitute, plus other benefits or resource savings that the consumer receives from obtaining the foreign-produced good. It is not entirely clear, however, which prices or values are actually used in working out foreign trade plans by Gosplan and the relevant ministries.¹³

At the same time, debate continues regarding the proper official internal exchange rate. There is still some aversion to setting this rate equal to the “marginal” rate ($e_s^{"}$) of the figure. The basic alternatives can be illustrated by focusing on exports. One approach would be to consider the ith exportable profitable only if its real implicit internal exchange rate (Z_xi/$P_{xi}^{\prime }$) is less than the weighted average real implicit exchange rate ($e_{xa}^{"}$) for, say, last period’s exports. Assume for simplicity that the valuta value of exports last period were equal to O$V_0^{\prime }$ in the figure. The average export rate, $e_{xa}^{"}$, will lie somewhere between A and $e_s^{"}$ on the vertical axis. Clearly, many exports with a higher real implicit exchange rate than $e_{xa}^{"}$ (up to $e_s^{"}$) would also be profitable. A more sophisticated “average” shadow exchange rate approach would limit exports to those products with (Z_xi/$P_{xi}^{\prime }$) less than the weighted average real rate for past imports ($e_{ma}^{"}$). As Gardner (1983) suggests, this so-called “import equivalent” criterion is an improvement over the average export rate, because it at least takes into account the (average) economic value to the country of obtaining foreign exchange. Assume again, for simplicity, initial imports with a valuta value of O$V_0^{\prime }$. Clearly the rate $e_{ma}^{"}$ will lie somewhere between C and $e_s^{"}$. Following this particular average-rate criterion, however, would in this case lead the planners to overexport (see Shagalov (1983), Chapter 3, and Gardner (1983)).

According to Zakharov and Shagalov (1982), Gosplan currently uses some twenty internal exchange rates, with different rates “for trade with individual CMEA countries, other socialist countries, and developed capitalist and developing states” (p. 1023). Although it has been observed by Shagalov and Faermark (1981) that the present Gosplan approach to calculating these rates cannot be considered optimal, as a practical matter these rates often may not differ significantly from the shadow rates because the planners in effect may calculate the official internal rate by splitting the difference between the average export and the average import rates.

The logic of a formal optimization approach to foreign trade, in which either domestic output or consumption is maximized, would be to permit structural shifts in the world economy to influence, through foreign trade, the structure of domestic production. Although the formal models developed by Soviet economists stress the desirability of integrating the foreign trade sector more closely into the national economy, few appear to propose that structural development within the Soviet Union should take place independently of a policy of relative self-sufficiency within the CMEA region. Although the planning of foreign trade on the basis of more meaningful internal exchange rates appears to have become official policy, foreign trade “effectiveness” is still listed in the 1980 Gosplan guidelines as only one of several important factors determining the foreign trade plan.

Zakharov and Shagalov (1982) have proposed setting domestic wholesale prices for exportables between a lower limit, equal to domestic costs of production (presumably the shadow price, Z_xi), and what we have called the calculative price ($P_{xi}^{"}$) derived from the shadow exchange rate. This is seen as a way to ensure that neither industry nor the foreign trade organizations have a financial interest in exporting goods that are “unprofitable” to the economy. It is suggested that setting domestic prices in this fashion would give to industry a financial interest to reduce costs and, to the foreign trade organizations, an incentive to maximize the foreign currency price received for exports. The implementation of such a proposal, however, would depend on the devolution of considerably greater authority for decision making in foreign trade to the industrial enterprises and the trade organizations.

In sum, it is evident that the foreign trade planners at Gosplan are increasingly using more economically meaningful internal exchange rates in planning the composition and direction of foreign trade. It is also clear, however, that Soviet foreign trade patterns also depend on a number of other criteria. There is some evidence that exchange rates have a price-determining role in some sectors of the Soviet economy, but this has not become a general tendency. As long as the foreign trade organizations and industrial enterprises have limited authority regarding foreign trade, it is likely that the official internal exchange rate will remain chiefly a planning device at the highest levels of the planning hierarchy. The present role of this exchange rate in macroeconomic stabilization or structural adjustment appears to be negligible.

V. The German Democratic Republic

The industrial combines, or Kombinate, have emerged since 1977–79 as the most important unit of economic management in the German Democratic Republic. Over the past few years, the Kombinate have in theory also taken on a greater share of responsibility for decisions on foreign trade. For many of the Kombinate, foreign trade is vital; of about 150, it is estimated that 90 sell over 20 percent of their output abroad, and about 40 of these have export quotas of 40 percent or more.

Foreign trade is actually carried out by specialized foreign trade enterprises, but the organizational relationships that these enterprises have with the Kombinate are varied. There appear to be essentially four arrangements: an enterprise may report directly to a Kombinat; it may report directly to the Ministry of Foreign Trade; it may report directly to this Ministry, but carry out foreign transactions for several Kombinate; or it may report directly to an industrial ministry but have several foreign trade departments, each of which is responsible both to the ministry and to a particular Kombinat belonging to that industry.¹⁴

With a view to strengthening the export interest of the industrial enterprises, the Kombinate, and the trade enterprises, export profitability has at least in theory entered into calculations of the “unified enterprise result” since 1971.¹⁵ These “results,” or profits, in turn affect contributions to the bonus funds. Export profitability involves comparing a calculative price with the domestic price ($P_{xi}^{"}$) inclusive of commissions of the foreign trade enterprises and other minor costs. The calculative price is equal to the foreign currency price times a full exchange rate (or Umrechnungskoeffizient), which is defined as the average domestic cost of earning a unit of foreign exchange, e_xa (Blessing, FrÖhlich, and Grote (1984)).¹⁶

This export profitability calculation is by no means meant as a rigid criterion for whether or not to export. Indeed, to use the average domestic cost of earning a unit of foreign exchange as a “cutoff” point would be self-defeating, as it would lead over time to ever lower levels of exports. It is rather meant to give the Kombinate and the trade enterprises a better sense of the relative profitability of exportables, and to encourage them to raise their average level of export profitability (Blessing, FrÖhlich, and Grote (1984)). Export profitability is only one of several criteria by which an enterprise’s performance is evaluated, and profitability per se still appears to be viewed as less important than the fulfillment of targets for trade volumes and foreign exchange earnings in the annual trade plans. Rather than focusing on the level of profits, evaluation appears to be directed more to comparisons of actual against planned profitability or to the growth of the enterprise’s average export profitability (Kupferschmidt, and others (1982)).

There is little evidence to suggest that export calculative prices are permitted to influence domestic wholesale prices directly. It is possible, however, that a Kombinat, faced with an increasing calculative price (as the result, say, of a rising world market price) might have a financial interest in obtaining above-plan allocations of domestic inputs by bidding up their prices. Whether the Kombinate are at present in a position to do this is, however, open to question. The role that world market prices should play in directly influencing the domestic prices of exportables is a widely debated issue in the German Democratic Republic (Ebersbach (1983)).

The major channel for the direct transmission of changes in world market prices appears to be imports. It has been suggested that domestic wholesale prices for virtually all products for which there are few close domestic substitutes are now based on transactional prices (Aufwandspreise). For many importables, however, such as certain raw materials and spare parts, these domestic prices are fixed for the duration of the Five-Year Plan. If close domestic substitutes are available, then the domestic prices of such importables are determined by domestic costs (Blessing and Grote (1982)).

The foreign trade enterprises and the Kombinate are expected to calculate import profitability as well. The standard for profitability is again the average domestic cost of earning a unit of foreign exchange (e_xa). In this case, imports are particularly encouraged that have a P_mi/$P_{mi}^{*}$ ratio greater than this average cost. Economists in the German Democratic Republic point out that such comparisons are meaningless when domestic prices are indeed set equal to transactional prices; the profitability calculation only makes sense if domestic substitutes exist with autonomous domestic prices. In this case, however, it has been suggested that the new Soviet guidelines for estimating the economic value of importables should be followed in establishing a domestic value for the comparison (Blessing and Grote (1982)). Economists in the German Democratic Republic also stress that for a country so dependent on raw material and fuel imports, it is vital to take full account of the real cost of the import content of exports in calculating export profitability.

If this is a realistic portrayal of how the foreign trade system works in practice in the German Democratic Republic, both changes in world (and intra-CMEA) prices and in the average domestic cost of earning foreign exchange might influence the balance of trade and structural adjustment within the country. The principal differences from the Soviet case, in theory at least, seem to be that there is more scope in the German Democratic Republic for the transmission of external price changes through the exchange rate, and a greater share of the responsibility for responding to changing world prices is borne by the industrial managers rather than the planners. Changes in world market relative prices may influence the domestic price structure of importables in the German Democratic Republic and the pattern of demand for imported versus domestically produced inputs. Through the export profitability calculations, the relative attractiveness to the Kombinate of different lines of production may also be altered.

Changes in the “average” exchange rate on which such profitability calculations are made in principle should affect the volume and structure of trade. This impact should not be exaggerated, however, because this exchange rate is itself a function of the level of trade (as well as domestic and foreign prices), and the level of trade is in turn very much influenced by the structural priorities and import decisions of the planners. The average nature of the official exchange rate also reduces its usefulness in achieving the “optimal” volume and composition of foreign trade. Finally, any changes in producer prices that do take place in this system have little direct impact on the structure and level of consumer prices. It is likely, however, that wholesale price changes have been passed through to the retail level, by administrative decision, more than the official retail price indices would suggest.

Some observers are quite skeptical about the degree of autonomy possessed by the Kombinate and about the role that the exchange rate system has played in determining the pattern of foreign trade and domestic adjustment.¹⁷ Economists in the German Democratic Republic have also recently written of the need to strengthen the interests of the Kombinate in foreign trade and for them to integrate their foreign trade enterprises more fully into their production and trade activities (Ebersbach (1983), and Kosser and Kupferschmidt (1983)).

VI. Poland in the 1970s

The theoretical foundations for the Polish exchange rate system of the 1970s were set out in a number of pioneering works by Polish economists in the 1960s, including, in particular, Trzeciakowski (1965). The basic approach, explained in detail in Trzeciakowski (1978), is essentially that used to determine the shadow exchange rate of the figure. In addition, Trzeciakowski and his collaborators were concerned with the development of various decomposition algorithms for calculating shadow prices and exchange rates, and also with the practical problem of maximizing the efficiency of foreign trade in an intentionally mixed system that combined direct and “indirect” management of foreign trade. Although Polish policymakers had come to accept the need for less centralized decision making in foreign trade, centrally determined import and export quotas were still to apply to a large share of trade.

The basic optimization approach was applied on an experimental basis in the late 1960s in some industrial branches, using calculative prices only. Industrial enterprises and foreign trade organizations were assigned calculated shadow prices for key imports and shadow internal exchange rates for each currency area. The calculative prices were not allowed to affect actual domestic prices, but the calculated profits of the enterprises involved in the scheme affected their bonus funds. At the same time, a truly “transactional” system (in which domestic prices were affected by transactional prices), was also experimented with in one industrial trust (Trzeciakowski (1978)).

The “economic maneuver” in Poland that began in 1971 was accompanied by a fairly significant experiment in the indirect management of foreign trade. Industrial ministries and the “large economic organizations” were permitted to trade on their own account or through the trade organizations on a commission basis.¹⁸ Although the large organizations were to have considerable autonomy, they were still given many value and trade volume quotas by the central authorities. The system that evolved, therefore, was one of mixed direct and indirect management of foreign trade.

Although the strikes in 1970 had led to the rescinding of various internal price reforms regarding consumer goods, most of the planned producer price changes were retained. Transactional prices designed to affect domestic prices were applied widely in exports. According to Plowiec (1973–74), negotiated domestic wholesale prices could now, potentially, encourage profit-oriented industrial enterprises to lower costs and raise product quality, while at the same time the trade organizations would have a direct interest in maximizing the foreign currency prices received for their exportables as well as buying from the lowest-cost domestic firms.

There was relatively little application of transactional pricing on imports (Plowiec (1973–74)). Although transactional prices were apparently applied to raw materials and machinery imports for which there were no domestic substitutes, these prices were fixed for the Five-Year Plan. In practice, most domestic raw material prices were not constructed on the basis of a uniform criterion, and in many cases failed to cover domestic costs of production (Bohm (1983)). While price equalization was terminated in principle, it must have remained very important on the import side and also would have had to apply to centrally directed exports.

The transactional prices that were established were based on an official internal exchange rate that was slightly higher than the average implicit internal exchange rate for exports. The dollar-area official internal rate of 15 zlotys (unpublished and meant to be known only to Polish planners and enterprises) was set at 110 percent of the estimated average implicit rate, which rendered approximately one third of existing exportables unprofitable from the enterprises’ standpoint.¹⁹ The official rate was based on an analysis of the 1967–68 cost structure of Polish exports, and was fixed for the period of the Five-Year Plan. Many foreign trade experts advocated setting the official internal rate at its marginal level, but largely because of a fear of inflationary consequences, the (approximate) average rate was adopted. According to Bohm (1983), the existing structure of domestic relative costs was such that energy carriers, which were frequently priced below domestic cost, and products intensive in raw materials were the most profitable exports. Exports that were “unprofitable” (given the prevailing implicit and official internal exchange rates), which included the products of light industry and many foodstuffs, had to be mandated by the central authorities and subsidized on a differentiated basis. This meant more intervention in the activities of enterprises than had originally been foreseen, and the automatic subsidies reduced the pressure on these exporting firms to lower their costs.

Widespread de facto price equalization on importables meant that the explosion in world prices of energy and raw materials in 1973–74 was not directly transmitted to the Polish economy.²⁰ At the same time, increasing world inflation and particularly rapid price increases for energy and manufactures intensive in raw materials were reflected in rising transactional prices for Polish exportables. The result was a rapid increase in effective protection for a wide range of processing industries, and quasi-autonomous enterprises were led to accelerate their imports of raw materials for processing and re-export (Bohm (1983) and Fallenbuchl (1980)).²¹ The authorities responded in 1975 by requiring exporters to compute their profits using shadow prices for imported inputs, and at the same time a windfall profits tax was imposed on exports. These enterprise responses to world market price developments could have been forestalled, of course, by a combination of a wider range of transactional pricing of imports and by a revaluation of the internal zloty relative to the external zloty.

At the beginning of the period of the 1976–80 Plan, the official internal exchange rate for the dollar area was indeed reduced from 15 to 13. Some domestic prices for raw materials were also raised. According to BÖhm (1983), these policy actions in effect reinstated the official internal rate to about 110 percent of the average implicit internal exchange rate. Over time, however, windfall profits on exports increased, exporters were once again required to use shadow prices, and a tax was imposed on the raw material value of exports. At the same time, to promote exports in response to the deteriorating balance of payments, branch-differentiated export rebates were paid to those large organizations or industrial associations that exported goods with an average implicit internal exchange rate greater than 87 percent of the official internal rate. Only coal and other energy carriers did not require export subsidies under this arrangement. These subsidies were mainly financed by higher taxes on the truly profitable exporters, so that the system was sliding back towards full price equalization (BÖhm (1983)).

Plowiec (1980–81) has suggested that a “submarginal” official exchange rate, which would make roughly 70-80 percent of exportables profitable to the enterprises, is probably satisfactory in a regime in which there is both centrally mandated and liberalized trade. This was also the recommendation of a special Commission for Foreign Trade Reform, and in 1980 the official internal exchange rate for the dollar area was accordingly raised to 14.5, which was about 30 percent higher than the estimated average implicit internal rate. The Commission also recommended that the official rate be changed if the average implicit rate on exports varied by more than 3 percent. In 1982 Poland began to quote publicly a full official commercial exchange rate.

A tantalizing empirical issue that emerges from the Polish experience is the extent to which the nature of the Polish foreign trade and exchange rate systems, the relative fixity of the official internal rate, and the structure of taxation and subsidization of trade, contributed to the deterioration of the Polish economy in the 1970s. It is possible that the initial partial decentralization of decision making on trade, combined with the overvaluation of the zloty and the policy of subsidizing raw material imports, added significantly to the economy’s structural problems and excess demand pressures. The Polish experience seems to highlight a serious issue for the authorities of a partially decentralized foreign trade system, namely the motivation and the ability of enterprise managers to use efficiently the modified environment to their own, but not necessarily the economy’s, best advantage. The importance of establishing “correct” financial parameters in such a modified system cannot be exaggerated. This is certainly appreciated by many Polish economists who continue to probe for ways to promote maximum efficiency in decentralized trade that is also consistent with the center’s need to plan the overall volume and to a large extent the geographical and commodity composition of foreign trade.²²

VII. Hungary

Of all the exchange rate systems of the CMEA countries, that of Hungary is probably best known. Much has been written about it since the introduction of the New Economic Mechanism in 1968. (See, for example, Brown and Marer (1973), Kozma (1981), Marer (1981b), and Balassa (1983)). One cornerstone of this new mechanism was the development of a foreign trade coefficient or multiplier that provided price links to foreign markets. For all practical purposes, this was a full exchange rate, although until 1976 it coexisted with what has been referred to in this paper as an official external rate. In 1976, the external rate was eliminated for purposes of calculating trade statistics, and the multiplier was thereafter referred to as the commercial rate. In 1981, the commercial and noncommercial exchange rates were unified into one official exchange rate.

Another essential facet of the New Economic Mechanism was the establishment of a more flexible price system with different categories of producer and consumer prices: “free” prices—although changes in these typically required consultation with the central price authorities; “limited flexibility” prices; and fixed prices. For a wide range of goods, price equalization was abolished, and for others it was intended to be internalized within the enterprises by means of a “reserve for smoothing out price differences.” Enterprises, now in principle released from the constraints of detailed central planning, were encouraged to increase profits, including those earned in foreign trade activities. Some larger enterprises or trusts were given foreign trade rights; in other cases foreign trade organizations were to engage in foreign trade for profit-oriented enterprises on a commission basis (Brown and Marer (1973), Kozma (1981), and Marer (1981b)).

The dollar- and ruble-area exchange rates established in 1968 were set approximately equal to the estimated average domestic cost of earning one unit of foreign exchange (e_xa) for the period 1960–64. A lengthy debate both preceded and followed this decision, with many reform-minded economists recommending that the official rate be set equal to the marginal domestic cost of earning a unit of foreign exchange (Brown and Marer (1973)).²³ According to Kozma (1981), the main considerations for the use of an “average” rate were to limit domestic inflation and to maintain “stable surroundings for the enterprises.” In effect, it was desired to “facilitate a gentle transition from the old system to the new,” and to maintain greater scope for continued central intervention (through, for instance, differentiated export subsidies). Such intervention permitted both the planners and enterprise managers to continue to think mainly in terms of quantities rather than values, and avoided significant dispersion of profit rates for exporting enterprises.

Although the profitability of enterprises was now directly affected by the transactional prices based on the exchange rate, the impact of changes in foreign currency prices on the domestic prices of nontraded exportables appears to have been only indirect after about 1973. By then, prime costs had become the predominant basis for increases in domestic exportable prices (Hare (1976)).²⁴ Prime costs could of course increase if world market prices for imported intermediates were not offset by subsidies. Moreover, higher transaction prices for exportables, even if not permitted to influence domestic prices directly, would in principle induce enterprises to allocate resources into the production of exportables, and this, under conditions of increasing marginal costs, would put upward pressure on the prime cost and thus the domestic wholesale price of exportables.

The “active” exchange rate revaluations of the mid- and late- 1970s were partly designed to nullify imported inflation; this suggests that transactional pricing was at least indirectly affecting the domestic prices of exportables and directly the prices of some importables (Brown and Tardos (1980)). Another motive for revaluation would have been to eliminate the windfall profits of exporters of goods intensive in raw materials and fuels who were benefiting from being able to import them at subsidized prices.²⁵

In principle, profit-oriented enterprises should have been sensitive to any changes in domestic relative prices or movements in transactional prices relative to domestic prices that were caused by changes in world market relative prices, the exchange rate, or both. Many Hungarian economists expressed skepticism in the late 1970s and early 1980s, however, that enterprises really reallocated resources in a significant way in response to such relative price changes. Kozma (1981) suggested that setting the official exchange rate at the “average” level made large-scale export subsidies inevitable and that this reduced the profit “interestedness” of enterprises and their sensitivity to changes in relative prices. Tardos (1980) argued that enterprise managers in Hungary were by no means short-run profit-maximizers, but interested above all in demonstrating monotonically increasing profits over time. This, too, would leave them relatively insensitive to price changes, at least in the short run; and in an environment of far-reaching export subsidization, their perception may have been that their time could be better spent bargaining for higher subsidies than shifting resources among different activities and investing in export-oriented lines of production. Kornai (1980) was also skeptical of the price sensitivity of enterprise managers that were only subject to what he called the “soft” budget constraint, and who realized the authorities’ reluctance to permit the bankruptcy of firms. Kornai (1982) also suggested that in any event the reallocation of, say, exportables to foreign markets in response to higher prices may be effectively neutralized by nonprice pressures coming from the authorities to alleviate or avoid excess demand pressures on domestic markets.²⁶

In the late 1970s, there was accordingly a lively debate over whether Hungary should pursue an active exchange rate policy, and if so, whether devaluation or revaluation of the forint would be better. Not surprisingly, those with an interest in promoting exports, such as the Ministry of Foreign Trade, or in reducing the budgetary burden of the export subsidies, such as the Ministry of Finance, tended to approve of devaluation. Others, concerned primarily about domestic price stability, were more interested in revaluing the forint. Those economists who perceived Hungary’s foreign trade elasticities to be low were skeptical that devaluation would seriously improve the deteriorating balance of payments. Kozma (1981) saw some potential for devaluation to encourage exports, but suggested that the demand for imports was very price inelastic because imports of investment goods were determined principally through the moral suasion of the central authorities. Portes (1979) considered Hungarian trade elasticities low with respect to domestic import demand and export supply, and also with respect to foreign demand for that country’s exports. The logic of Wolf’s model of a modified CPE (1978), on the other hand, was that if price elasticities were more than negligible but mainly only domestic prices of exportables were affected by devaluation, an increase in the exchange rate could worsen the trade balance in foreign currency terms.

This price system was replaced in 1980–81 by the “competitive” system. Now the domestic prices of imported intermediates were to be tied directly to the tariff-inclusive transactional prices. Energy prices, however, were still fixed centrally and were only modified “intermittently” as world market prices changed (Balassa (1983)). The “competitive” rules on the export side, however, were more complex. While wanting domestic firms to be sensitive to changes in world market prices for exportables, the authorities also sought to limit the domestic inflationary impact of such changes. This was especially critical for an economy in which most industrial branches had ended up with monopolistic market structures after a quarter century of central planning.

Under this competitive system, enterprises that exported more than 5 percent of output for convertible currency (accounting for about two thirds of Hungarian industry) were subject to three basic rules: (1) the average price of their domestically marketed output could not rise faster than the average increase of their transaction export prices (including a rebate for imputed indirect taxes); (2) the average profit margin on domestic sales could not exceed that on exports; and (3) individual product prices on domestically marketed goods could not exceed the actual or hypothetical convertible currency (transactional) import price for those particular products.

This competitive price system was the subject of various criticisms. Balassa (1983) noted, for example, that firms might raise their convertible currency prices to increase their overall rate of export profitability, permitting higher domestic profits, but that if they faced elastic export demand this would lead to reduced export revenue and losses for the national economy. It was also suggested that enterprises might seek to eliminate less profitable exports altogether so as to raise their overall level of export profitability (and again raise the scope for raising domestic prices). This would contribute to a further decline in export revenue, and could constitute a loss to the national economy if these exports that appeared unprofitable (because of the below-marginal exchange rate) were indeed socially profitable. Other firms might be tempted to cut back exports so as to be exempt from the “5 percent” rule that made them directly subject to the “competitive” rules.

These and other problems led to changes in price-building regulations in 1981–82, and in 1984 further modifications were introduced. The number of firms to be covered by the competitive rules is being increased, and certain enterprises are now subject only to the third rule noted above. Enterprises accounting for as much as 35 percent to 40 percent of manufacturing output may now be able to qualify for this more liberal application of the competitive rules, but a demonstration of “responsible” use of domestic monopoly power will be a criterion for continuing to qualify.

Given the greater scope for the official exchange rate to influence domestic prices in Hungary, compared with the other CMEA countries reviewed here, to what extent has the Hungarian exchange rate system affected the process of structural adjustment and macroeconomic balance? Regarding structural adjustment, van Brabant (1977), in a detailed study covering 1969–75, found little systematic evidence that the Hungarian export structure, in trade with either the dollar or ruble areas, had shifted away from branches with relatively high domestic costs of earning foreign exchange into branches with relatively low costs. Simon (1984), in a comprehensive empirical study for the 1965–79 period, found that over 40 percent of Hungarian branches experienced a negative (instead of the expected positive) correlation between changes in their share of total exports and changes in their ratio of transactional prices to domestic costs.

Regarding the impact of devaluation on the trade balance, Balassa (1983) suggests that much of the previously mentioned skepticism may be unwarranted. Marer (forthcoming) suggests, however, that under the “competitive” system, a devaluation in general is unlikely to have a significant effect on the convertible currency trade balance. This is mainly because many domestic prices (through competitive pricing) and CMEA forint trade prices (through a system of producers’ differential turnover taxes and subsidies) are now linked fairly directly to transaction prices on exports to the convertible currency area. A devaluation, it is argued, will therefore cause very little, if any, change in the hard currency transactional price relative to these other prices, and there will be correspondingly little incentive for producers to reallocate resources into production for the convertible currency area. This, it should be noted, is in contrast to the stylized market economy case in which the domestic prices of exportables ultimately rise precisely because of excess demand pressures caused by the reallocation of resources into exports as a result of the initial devaluation. Marer’s point is that the Hungarian “competitive” system, by only emulating the outcome but not the process of the competitive free market, may preclude the allocational impact that characterizes the latter.

It should be noted, however, that administratively determined adjustments in transaction prices for the ruble area may only follow changes in those for the dollar area with a lag, although it is doubtful in any event that short-run changes in relative ruble-/ dollar-area transactional prices would lead to resource reallocation. More important, not all domestic prices are subject to the “competitive” system, and in any event, enterprises are by no means compelled to raise domestic prices pari passu with convertible-currency export transaction prices. If demand management policies were sufficiently restrictive, producers might well not raise domestic prices until the impact of devaluation on resource allocation had itself created domestic excess demand pressures. Moreover, even if a change in the exchange rate were to have little effect, the emerging domestic price-building role of the transactional prices in this system, together with the enterprise’s presumed interest in raising its profitability, might give the firm the incentive to concentrate investments in those exportables expected to enjoy relatively rapid world market price increases in the future.

Marer (forthcoming) also suggests that because Hungary is so dependent on imported raw materials and intermediates, and because these are now priced domestically mainly on a transactional basis, a devaluation will tend to raise the domestic material costs of production virtually in proportion to the increase in the price of exportables. This effect, too, might be expected to diminish, but not necessarily to eliminate, the impact of a devaluation. Observe, however, that this argument would apply to any country that imports most of its raw materials, whether it is a modified CPE or a market economy.

It has been suggested that because the nominal wage is directly controlled by the authorities in Hungary, the expenditure-reducing impact of devaluation is perceived by the authorities to be less important than it might be by the government of a market economy. Under the more parametric wage-regulation system now in effect, however, and given the increasing importance of incomes earned outside the socialized sector, it is not obvious that the Hungarian authorities can control nominal incomes as precisely as commonly believed. If Hungarian workers are subject to money illusion too, they may more readily accept a decline in the real wage through price increases rather than by a reduction in the money wage, and the exchange rate, which now in principle affects a wide range of prices, could be seen as a relatively efficient means to change the price level. Finally, the separate effect that a devaluation and attendant price increases would have on real money balances, and therefore possibly on real expenditure, should not be ignored. If relative price changes are indeed minimized by the “competitive” system, or if many of the aforementioned alleged obstacles to responsiveness by Hungarian enterprises to relative price changes persist, the expenditure-reducing impact of devaluation might dominate the substitution effect.

Many economists have suggested that together with tighter financial discipline, a further liberalization of imports would raise the effectiveness of the exchange rate system and exchange rate policy in Hungary.²⁷ Liberalized imports would put downward pressure on domestic prices of import substitutes. This might reduce the bias towards import substitution, and by increasing de facto competition domestically, remove one of the main reasons for preserving the artificial “competitive” pricing system. This in turn might make the exchange rate a more effective policy instrument.

VIII. Conclusions

This paper has presented a uniform terminology and common analytical framework for comparing the exchange rate systems of stylized market economies, classical CPEs, and MPEs. The exchange rate systems and policies of four CMEA planned economies (the U.S.S.R., the German Democratic Republic, Poland— in the 1970s, and Hungary), were then examined.

These four countries represent within the CMEA a spectrum of the extent to which decisions on foreign trade have been decentralized and the exchange rate has taken on a direct role in determining prices. To the extent that this decentralization has occurred, the exchange rate has in principle a broader role to play in both economic stabilization and structural adjustment. The paper analyzes this role in these four planned economies.

The authorities of the Soviet Union are increasingly relying on more economically meaningful exchange rates, along with other criteria, to plan foreign trade. Exchange rates may have a function in determining prices in some sectors, but not as yet a general function in this regard. As long as foreign trade organizations and industrial enterprises have severely limited autonomy, the exchange rate will remain chiefly a planning device at the highest levels of the planning hierarchy.

Large industrial Kombinate in the German Democratic Republic have somewhat more interest in, and autonomy over, foreign trade than do their industrial counterparts in the Soviet Union. The exchange rate influences the domestic price of some imports, and it may also have a limited influence in resource allocation, at the level of the combine, for exports. The function of signalling prices, and therefore the potential use of the exchange rate in structural adjustment, may eventually be enhanced in that economy.

The Polish exchange rate system and policies during the 1970s are instructive. Initially, an economically meaningful, though overvalued, exchange rate was established, and foreign trade was partially decentralized. Continuing and growing price distortions, however, together with growing excess demand pressures, led to a renewed increase in state intervention in foreign trade. This highlights some of the dangers involved in a partially liberalized foreign trade system responding to distorted financial parameters. The Polish experience raises the question whether these distortions may have added to the structural and macroeconomic problems that evolved during the 1970s.

Experimentation with determining domestic prices according to exchange rates and with liberalizing foreign trade has been most striking in Hungary. The potential significance of the exchange rate as a transmitter of signals from the world market and as an instrument of economic stabilization is much enhanced in the present Hungarian context. Past experience raises some doubt about the actual economic impact of exchange rate changes in Hungary, but as yet there has been insufficient theoretical and empirical analysis of the effects of the exchange rate in this economy to establish firm conclusions.

This study should demonstrate that we must go beyond the very stylized models (of both market economies and CPEs) in attempting to understand the actual impact of exchange rates in planned economies. The paper also shows, in the author’s view, one of the values of comparative analysis—the diversity of exchange rate systems documented here may prove useful in analyzing and speculating on the potential role that exchange rates might play in different planned economies.