Economic Structure, the Exchange Rate, and Adjustment in the Federal Republic of Germany: A General Equilibrium Approach
Author: THOMAS MAYER

The effects of an appreciation of the deutsche mark are traced with the help of a computable general equilibrium model under alternative structural policy scenarios. In the first scenario, characterized by severe structural rigidities, the contractionary effects of exchange rate appreciation dominate the expansionary effects, so that gross domestic product and employment fall and the external surplus declines only slightly. In the alternative case of free movement of goods, services, and factors, the expansionary effects of the appreciation become more prominent as supply and demand respond much more readily to the relative price changes. [JEL 431]

Abstract

The effects of an appreciation of the deutsche mark are traced with the help of a computable general equilibrium model under alternative structural policy scenarios. In the first scenario, characterized by severe structural rigidities, the contractionary effects of exchange rate appreciation dominate the expansionary effects, so that gross domestic product and employment fall and the external surplus declines only slightly. In the alternative case of free movement of goods, services, and factors, the expansionary effects of the appreciation become more prominent as supply and demand respond much more readily to the relative price changes. [JEL 431]

The effects of an appreciation of the deutsche mark are traced with the help of a computable general equilibrium model under alternative structural policy scenarios. In the first scenario, characterized by severe structural rigidities, the contractionary effects of exchange rate appreciation dominate the expansionary effects, so that gross domestic product and employment fall and the external surplus declines only slightly. In the alternative case of free movement of goods, services, and factors, the expansionary effects of the appreciation become more prominent as supply and demand respond much more readily to the relative price changes. [JEL 431]

When efforts to redress the international economic imbalances among major industrial countries began in 1985, the emphasis was on macroeconomic policies and exchange rate changes. With the room for maneuver in these policy areas diminishing and the adjustment being far from complete, however, the focus has recently shifted to structural policies. With regard to the Federal Republic of Germany in particular, the view has gained strength that trade liberalization as well as deregulation of goods, services, and labor markets would make an important contribution to the external adjustment of this economy by increasing its responsiveness to the exchange rate changes that have already taken place.

The present paper seeks to examine this hypothesis by investigating the effects of structural rigidities and protectionist practices on the adjustment process in Germany. The analysis is conducted in the form of an illustrative quantitative exercise in which the effects of an exogenous appreciation of the deutsche mark are examined under different structural policies. The results of this investigation indicate that structural policy action, supported by macroeconomic policies consistent with the nominal targets of the German authorities’ medium-term economic strategy, would indeed facilitate external adjustment in an environment of higher growth, lower unemployment, and lower inflation. In the following section, several protectionist practices and rigidities that could be obstacles to adjustment are identified. Section II outlines an analytical framework for a quantitative analysis of the effects of these policies and rigidities on the German economy, and Section III presents the results of the illustrative exercise. Section IV discusses possible costs of adjustment to more liberal policies, and some concluding observations are made in Section V.

I. Impediments to Adjustment

In an open economy that is free from restrictions on international trade and the domestic flow of resources and in which foreign currency prices of traded goods are largely determined on world markets, an appreciation of the exchange rate lowers the domestic-currency price of traded goods and, everything else being equal, also the general price level. The decline in prices elicits higher real consumption (the real balance effect), leading to higher domestic demand and to an increase in imports. At the same time, the supply of exports and import substitutes declines in response to lower domestic-currency prices of these goods. As a result the volume of exports falls relative to that of imports. Higher domestic demand militates against a fall in the price of nontraded goods (the supply of which is limited by the production potential of the economy), even though there might he downward pressure on input costs: thus the increase in the relative price of these goods that was elicited by the decline in the price of traded goods is reinforced. Consequently, there is a shift of resources from the traded-goods industries to the sector producing nontraded goods. Real gross national product (GNP)—and, if there is slack in the labor market, employment—rises if the nontraded goods sector absorbs more resources than those released by the traded- goods sector.1 Then, the increase in domestic demand (weighted by its share in GNP) will exceed the sum of the increase in imports and the decrease in exports (each also weighted by its respective share in GNP).2

The smooth adjustment of the German economy to an appreciation of the deutsche mark, however, is likely to be hampered by several sectoral policies and rigidities. First, the system of variable subsidies and import levies in the agricultural sector and in the coal-mining industry is designed so that, in addition to providing considerable protection against external competition, it offsets the price effects of a currency appreciation for these sectors (Table 1). As a result, output and employment in these sectors are not reduced by the appreciation. In addition, a given nominal appreciation of the deutsche mark has a smaller effect on the domestic price level, domestic demand, and the resource allocation in the economy. Thus, for a given nominal appreciation, the real exchange rate appreciates more—so that exporters in the unprotected industries face a sharper decline in the external demand for their products (or a greater squeeze on their profits). In other words, unprotected exporters have to carry a larger burden of the adjustment. Whether the deterioration of the real trade position is smaller than in the absence of these protective policies depends on the size of net exports in the protected sectors relative to those in the unprotected sectors and on the price elasticities of export and import demand.

Table 1.

Protection in Selected Economic Sectors of the Federal Republic of Germany, 1982

article image
Sources: Weiss (1985), OECD (1985), and author’s estimates.

Calculated as the sum of the rate of effective protection and the ratio of subsidies to value added.

Second, import quotas, voluntary export restraints, and other non- tariff barriers impede the response of imports of various products (such as textiles, clothing, iron, steel, and ships) to an appreciation of the deutsche mark. Given the lesser competitive pressure from abroad in these sectors, the domestic prices of the goods are likely to decline less than they would in a regime of free trade. As described above, the results are smaller effects of the nominal appreciation on the general price level, domestic demand, and the reallocation of resources in the economy and a larger real appreciation and shifting of the burden of adjustment from import-competing to exporting companies and from the protected to the unprotected industries. In those industries that enjoy protection through subsidies, domestic prices have to follow domestic-currency prices of imported goods to maintain a constant share of imports in the domestic market. This linkage necessitates larger government expenditures on subsidies and thus requires either more government borrowing or measures to increase revenue. Increased government borrowing may contribute to higher interest rates (and, possibly, to a further appreciation of the currency) and, not unlike an increase in taxes, may raise the costs and reduce the competitiveness of the unprotected industries.

Third, the immobility of labor among industries and the numerous regulations that exist in the service sector (Table 2) might lower the supply response of the nontraded-goods sector to an improvement in the relative price of goods produced in this sector.3 As a consequence, prices of nontraded goods are likely to rise by more than they would have otherwise, resulting in less of a fall in the domestic price level, a smaller increase in domestic demand, a larger real currency appreciation, and, probably, a greater squeeze on the traded-goods sector.

Table 2.

Industrial and Sectoral Rules and Regulations in Germany, Mid-1980s

article image
article image
Source: Soltwedel and others (1986).

The importance of the industries affected by these protectionist practices, regulations, and rigidities for the German economy can be seen from Table 3. In the table agriculture and coal mining are consolidated into a sector producing “basic goods.”4 Iron, steel, textiles, clothing, and shipbuilding are aggregated into a sector dubbed “protected goods,” and industries producing services that are less likely to be traded internationally, such as professional services, retail trade, and the like, are combined in a “nontraded-goods” sector. The remaining industries are aggregated into a “traded-goods” sector.5

Table 3.

Stylized Structural Features of the German Economy, 1982

article image
Source: Statistisches Bundesamt (1987) and author’s calculations.

The German input-output table has a “government sector” that creates the goods used by government. This sector was included in the nontraded-goods sector.

Some of the industries included in the nontraded-goods sector do export and import. For the following analysis, however, trade of these industries was disregarded.

Excludes those self-employed in agriculture.

Estimate.

In 1982 the industries most heavily protected from international competition (that is, industries producing basic goods and protected goods) accounted for about 9.5 percent of gross domestic product (GDP) and employment. Basic and protected goods held a share in total exports and imports of 15 percent and 25 percent, respectively, and imports accounted for 17 percent of the total use of these goods—in the traded- goods sector, the import penetration ratio was 14 percent. Within the nontraded-goods sector, services subject to a significant degree of government regulation (Table 2) accounted for about 55 percent of value added.6 Notably, the return to productive factors was lower in the protected industries than in the rest of the economy. Differences in the return to labor were small relative to those in the return to capital, however, reflecting the strength and centralized organization of German trade unions. The basic-goods industries, with a relatively low return to capital, had the highest capital-labor ratio, whereas the nontraded-goods industries, which had the highest return to capital, were the most labor- intensive.

The nature and pervasiveness of certain protectionist practices and structural rigidities in the German economy would suggest that there are indeed impediments to the response of the economy to exchange rate changes. It is, however, impossible to give a quantitative assessment of the importance of these impediments without a numerically specified model that facilitates a “counterfactual” simulation—that is, the reaction of the German economy to an appreciation of the deutsche mark without the sectoral policies and structural rigidities that hinder the adjustment. An analytical framework that allows counterfactual simulations is presented in the following section.

II. Analytical Framework

In general, econometric models of an economy rely on historical evidence for the quantification of relationships among aggregated economic variables. Because these estimates depend on a given set of structural features, they are not suited to establishing counterfactual reactions of an economy to an exogenous shock, such as a currency appreciation, under different structural policy regimes. For this, a computable general equilibrium model of the sort described below is a better instrument. The advantage of computable general equilibrium models is that they are firmly based on microeconomic theory and that many parameter estimates are derived at the microeconomic level; such models are therefore likely to be more stable and less susceptible to the problem of “structural breaks.”

The computable general equilibrium model used for the following illustrative exercise is comparative static and belongs to the type of so- called Johansen models; it follows closely the version developed by Dixon and others (1982). The model emphasizes the role of relative prices and substitution possibilities in explaining trade flows and the commodity composition of domestic activity. The essential postulates governing producer and consumer behavior are profit and utility maximization. The model distinguishes the four productive sectors described in the previous section (basic goods, protected goods, traded goods, and nontraded goods), four types of final demand (investment, government consumption, private consumption, and exports) that are satisfied either from domestic sources or from imports, and three types of primary inputs (labor, capital, and land). It is numerically specified using a 1982 input-output table for Germany and parameter estimates culled from the literature. The equations of the model can be grouped into equations for input demand, final demand, supply behavior, market equilibrium, with several miscellaneous equations defining macroeconomic aggregates.

Input Demand

Producers are assumed to minimize the costs of production subject to a two-level production function. The first level imposes constant returns to scale and Leontieff complementarity between different types of intermediate inputs and between intermediate and primary inputs. The second level allows for constant elasticity of substitution between imported and domestically produced intermediate inputs and between different types of primary factors.7 The solution of this optimization problem yields a set of equations for producers’ factor demand. Thus, in a given sector (sector j), demand for both domestically produced and imported intermediate inputs (INTERINPUT) depends on the activity level and the prices of imports and domestic products (P) (a full set of variable names is given in Table 4):

Table 4.

Notation of Variables for the German Model

article image
article image
INTERINPUT(DOM,i,j)=F1[Z(j),P(DOM,i),P(IMP,i)],(1)(12)

where i = 1, 2, 3 and j = 1,…,4;

INTERINPUT(IMP,i,j)=F2[Z(j),P(DOM,i),P(IMP,i)],(13)(24)

where i = 1, 2, 3 and j = 1,…,4; and

INTERINPUT(D0M,4,j)=F3[Z(j)],(25)(28)

where j = 1, 2, 3, 4.

Equations (25)–(28) relate input demand for nontraded goods (which are produced by sector 4) only to activity levels in the demanding industries, since there are no imports to substitute for domestic deliveries.

The equations for primary input demand (labor, capital, and land) relate each sector’s demand to its activity level and to the prices of labor, capital, and land:8

LAB(j)=F4[Z(j),P(LAB),P(CAP),P(LAND)],(29)(32)

where j = 1,…,4;

CAP(j)=F5[Z(j),P(LAB),P(CAP),P(LAND)],(33)(36)

where j = 1,…,4; and

LAND(j)=F6[Z(j),P(LAB),P(CAP),P(LAND)],(37)

where j = 1.

Final Demand

In what follows, final demand is divided into four components: investment, household consumption, government consumption, and exports.

Investment

Investors are assumed to minimize costs subject to a given level of investment. Thus, the model allows for substitution between imports and domestically produced investment goods when relative prices change. Total investment in real terms is derived from total nominal investment [NINV (TOTAL)] and the price index for investment goods (IPI):

INV(DOM,i)=F7[INV(TOTAL),P(DOM,i),P(IMP,i)],(38)(41)

where i = 1,…,4;

INV(IMP,i)=F8[INV(TOTAL),P(DOM,i),P(IMP,i)],(42)(44)

where i = 1,…,3; and

INV(TOTAL)=NINV(TOTAL)/IPI.(45)

Household Consumption

Consumers are assumed to maximize their utility subject to an overall budget constraint. The resultant set of equations relates consumption demand for domestic and imported goods (CONS) to aggregate consumption and the prices of domestic and imported goods. Real aggregate consumption is derived from total nominal consumption [NCONS (TOTAL)] and the price index for consumer goods (CPI):

CONS(DOM,i)=F10[NCONS(TOTAL),P(DOM,1),P(DOM,2),P(DOM,3),P(DOM,4),P(IMP,1),P(IMP,2),P(IMP,3)],(46)(49)

where i = 1,…,4;

CONS(IMP,i)=F11[NCONS(TOTAL),P(DOM,1),P(DOM,2),P(DOM,3),P(DOM,4),P(IMP,1),P(IMP,2),P(IMP,3)],(50)(52)

where i = 1, 2, 3; and

CONS(TOTAL)=NCONS(TOTAL)/CPI.(53)

Government Consumption

Like private consumption, real government consumption is derived from nominal government consumption (NGOV) and the price index for goods consumed by the government. Because government services are solely provided by the “nontraded-services sector” (sector 4),9 this relationship can he expressed as follows:

GOV(DOM,4)=NGOV(DOM,4)/P(DOM,4).(54)

Exports

The model describes total export demand as a function of world prices for the exported good, measured in foreign currency:10

EX(i)=F14[FCP(EX,i)],(55)(57)

where i = 1,…,3.

Supply

The supply side of the model treats imports and domestic production.

Imports

As in the case of exports, the import supply of each good rises in response to an increase in the foreign currency price of imports:11

IMP(i)=F15[FCP(IMP,i)],(58)(60)

where i = 1,…,3.

Domestic Supply

Constant returns to scale and zero pure profits are assumed for all productive sectors.12 Thus, the individual producer chooses the output level that equates marginal costs with the output price:

P(DOM,i)=F16[P(DOM,j),P(IMP,w),P(LAB),P(CAP),P(LAND)],(61)(64)

where j = 1,…,4 but j#i; w = 1,…,3; and i = 1,…,4. Here, F16 is the marginal cost of production.

Domestic and World Market Prices

Domestic and import prices for traded goods are determined by the respective foreign currency prices of those goods, the exchange rate, the markups in foreign trade (for transport, wholesale, retail services, and so forth), the subsidy ratio (defined as unity plus the ad valorem rate of subsidies) in the case of domestic goods, and the duty ratio (defined as unity plus the ad valorem rate of import protection) in the case of imports:

P(DOM,i)=FCP(EX,i).XRATE.SUBSIDYRATIO(i).1/MARKUP(i),(65)(67)

where i = 1,…,3; and

P(IMP,i)=FCP(IMP,i).XRATE.DUTYRATIO(i).MARKUP(i),(68)(70)

where i = 1,…,3.

Market Clearing

The model is closed by a set of equations that links demand and supply by imposing market clearing on factor and product markets.

Markets for Primary Factors

The clearing of factor markets requires that the sum of the demand for labor and capital equals total supply.13 The supply of land equals demand for this factor by the sector producing basic goods:

LAB=ΣjLAB(j),(71)

where j = 1,…,4;

CAP=ΣjCAP(j),(72)

where j = 1,…,4; and

LAND=LAND(1).(73)

Domestic Production

The clearing of the product markets for output from the four sectors requires that total domestic production satisfy aggregate demand:

Z(i) =ΣjIMTERINPUT(DOM,i.j)+INV(DOM,i)+CONS(DOM,i)+GOV(DOM,i)+EX(i),(74)(77)

where i, j = 1,…,4.

Miscellaneous Equations

The model also includes equations that define macroeconomic aggregates, such as GDP, aggregate exports and imports, and the like (definitions of all variables, as noted earlier, appear in Table 4):

GDP=CONS(TOTAL)+INV(TOTAL)+GOV(DOM,4)+EM(78)
E=ΣiEX(i);i=1,,3(79)
M=ΣiIMP(i);i=1,,3(80)
Y=F81(EP/MP,GDP)(81)
EP=Σiw(x,i)FCP(EX,i);i=1,,3(82)
MP=Σiw(m,i)FCP(IMP,i);i=1,,3(83)
CPI=Σiw(cd,i)P(DOM,i)+Σjw(cm,j)P(IMP,j);i=1,,4;j=1,,3(84)
IPI=Σiw(id,i)P(DOM,i)+Σjw(im,j)P(IMP,j);i=1,,4;j=1,,3(85)
NCONS=F86(NABS)(86)
NINV=F87(NABS)(87)
NGOV=F88(NABS)(88)
TB=EP.EMP.M(89)
P(LAB)R=P(LAB)/CPI.(90)

Equation (81), for example, defines real income as a function of real GDP and the terms of trade. Equations (86)–(88) relate nominal private and public consumption and investment to nominal domestic absorption.14 Equation (89) defines the real trade balance as the difference between exports and imports, corrected for terms of trade effects. Equation (90) defines the real wage rate as the nominal wage rate deflated by the consumer price index.

Parameter Settings

Most of the parameters needed for the numerical specification of the model can be derived from the 1982 German input-output table (see Dixon and others (1982) for details). There are, however, several other parameters that have to be taken from econometric studies or, if not available, assumed. Table 5 shows the values of these parameters that were used for the simulations that follow.

Table 5.

Key Parameter Settings for the German Model

article image

Based on Lluch, Powell, and Williams (1977, p. 54).

Based on Lächler (1985, p. 85) and Fund staff calculations. These elasticities are assumed to be the same for all uses.

A simple Cobb-Douglas production function was assumed to characterize the protected-, traded-, and nontraded-goods sectors, whereas the substitution elasticity for primary factors in the basic-goods sector was set at a level that brought the output supply price elasticity in line with estimates from the literature.

The parameters give the percentage change of import supply in the respective sector in response to a 1 percent change in the foreign currency price of imports. These parameters enter equations (58)–(60) of the text.

The parameters give the percentage change of export demand in the respective sector in response to a 1 percent change in the foreign currency price of exports. These parameters enter equations (55)–(57) of the text.

The consumption parameters were derived from a linear expenditure system. The underlying household utility function was assumed to be additive; thus the uncompensated own-price elasticities (nii) and cross-price elasticities (nij, for i≠j) can be derived as follows (see Lluch, Powell, and Williams (1977)):

nii=(ξi/w)ξiαi[1+(ξi/w)]
njj=ξiαj[1+(ξj/w)],forij.

Here ξi represents the expenditure elasticity and αi the budget share for product i, whereas w is the Frisch parameter. The Frisch parameter was set at –1.83.15

The elasticities of substitution between domestically produced and imported products were specified on the basis of an econometric study for Germany by Lächler (1985). On the supply side, a Cobb-Douglas production function was assumed for the protected-, traded-, and nontraded-goods sectors, and a constant elasticity of substitution function was assumed for the basic-goods sector; the elasticity of substitution between primary factors in the latter sector was specified such that the supply elasticity was broadly in line with estimates from the literature.

No sector-specific information was found in the literature for the price elasticities of import supply and export demand. Appropriate values for these parameters were therefore determined by numerous sensitivity analyses with the model.

Solving the Model

The equations of the model are first transformed into a log-linear form that allows solving by simple matrix methods.16 The model in the form described above consists of 107 variables and 90 equations. Thus, to arrive at a solution, 17 variables have to be assumed exogenously. The selection of the exogenous variables and the design of the simulations are described in the following section.

III. Design of Experiments and Simulation Results

The selection of the exogenous variables of the model follows from the assumptions about the macroeconomic environment in which the exchange rate shock takes place and from the specific protectionist practices and structural rigidities that are to he analyzed. Although the macroeconomic environment was assumed to be the same for all experiments that were conducted, several different sectoral policies and rigidities were investigated that required changes in the set of exogenous variables. Table 6 provides an overview of the variables assumed exogenously in each of the simulations.

Table 6.

Values of Exogenous Variables in the German Model

(In percentage changes from base period)

article image
Note: The notation “(en)” indicates that the variable is endogenous in this experiment.

The simulations were intended to trace the short run (one- to two-year period) effects of a 10 percent increase in the nominal value of the deutsche mark (induced exogenously by a shift in international portfolio preferences) on the German economy. Thus, it was assumed that (1) capital and land are fixed factors of production in each sector, and investment does not add to the productive capital stock in an industry during the period under investigation; (2) proportional markups in foreign trade are constant; (3) real wages are constant and above full employment levels:17 and (4) fiscal and monetary policy continue to follow the nominal targets of the authorities’ medium-term economic strategy so that nominal domestic absorption is maintained unchanged. The last assumption implies that monetary policy supports nominal private consumption and investment by allowing monetary expansion to exceed nominal GDP growth when nominal GDP growth is reduced by a decline in the external surplus,18 and that the government maintains growth of nominal expenditures unchanged even if, because of lower nominal income growth, direct tax revenues decline and the budget deficit widens.

The baseline simulation establishes the counterfactual reaction of the economy to the appreciation of the deutsche mark—that is, the likely developments in the absence of protectionist practices and structural rigidities. As shown in Table 6, in addition to the variables mentioned above, the subsidy and duty ratios were assumed to be unchanged for this simulation. Four alternative scenarios were designed to illustrate the effects of the protectionist practices and structural rigidities. In scenario 1, it was assumed that variable import levies and export subsidies isolate the basic-goods sector from the effects of the appreciation. Thus, the prices of the domestically produced and imported products of this sector were assumed to remain unchanged while the subsidy and duty ratios were endogenously determined by the model so as to offset the change in the exchange rate.19 In scenario 2, it was assumed that quotas, voluntary export restraints, and other nontariff barriers present in the protected-goods sector prevent imports that compete with the goods produced by this sector from reacting to the appreciation.20 Hence imports of these goods were assumed to remain unchanged while the duty ratio, which in this case measures the tariff equivalent of the nontariff barriers in this sector, was endogenously determined by the model. In scenario 3, it was assumed that government regulations and labor market rigidities prevent output and employment in the nontraded-goods sector from reacting to the appreciation. This, of course, is an extreme assumption that serves only as a benchmark for a more complex reality. However, the regulations on market access and competition in industries such as retail trade, financial services, and professional services, and the (voluntary and systematic) restrictions to labor mobility all contribute to a sluggish supply response of the nontraded-goods sector to an increase in demand. Hence changes in the demand for goods produced by this sector lead to changes in the return to the productive factors employed there. By assumption, changes in the return to labor employed in the nontraded-goods sector affect the general wage level.21 Thus, employment is assumed to remain unchanged in the nontraded-goods sector while the real wage rate is determined endogenously in the model. Finally, in scenario 4, all of the above protectionist practices and structural rigidities together were assumed to influence the adjustment of the economy to the exchange rate shock.

The results of the baseline simulation and deviations from the baseline under the four alternative scenarios are presented in Tables 7 and 8. In interpreting these tables it is important to keep in mind that the results are contingent on the numerous model assumptions and the assumed parameter values. They are therefore more of the nature of controlled laboratory experiments than projections of actual developments of the German economy in response to an appreciation of the deutsche mark. The simulation results of the baseline scenario indicate the percentage deviation of a variable, in response to the exchange rate shock, from the level it would have otherwise attained after all the domestic and international effects have worked their way through the economy. The results of the alternative policy simulations are reported as percentage deviations of the endogenous variables from the baseline simulation.

Table 7.

Alternative Patterns of Adjustment to a 10 Percent Appreciation of the Exchange Rate: Macroeconomic Resultsa

(In percentage changes from base period or baseline)

article image

Results for the baseline simulation indicate percentage deviations of endogenous variables from the values they would have attained in the absence of the exchange rate shock. Results for the alternative policy simulations indicate percentage changes from the baseline simulation.

Absolute change in billions of 1982 deutsche mark.

Table 8.

Alternative Patterns of Adjustment to a 10 Percent Appreciation of the Exchange Rate: Sectoral Results

(In percentage changes from base period or baseline)a

article image

Results for the baseline simulation indicate percentage deviations of endogenous variables from the values they would have attained in the absence of the exchange rate shock. Results for the alternative policy simulations indicate percentage changes from the baseline simulation.

In the absence of protection and structural rigidities, a 10 percent appreciation of the exchange rate lowers domestic consumer prices by about 4.4 percent (Table 7). With nominal domestic absorption maintained by macroeconomic policies, the decline in the consumer price level leads to a corresponding increase in real domestic absorption.22 Import volumes are buoyed by higher domestic demand and the decline of the domestic currency price of imported goods relative to that of domestic products. The increase in imports is, however, smaller than that of domestic absorption because demand for imported inputs, which comprise about half of total imports, rises less than demand for imported consumer and investment goods.23 Exports, in contrast, decline by around 5.5 percent because of the increase in prices expressed in foreign currency.24 As a result, the real trade balance deteriorates by about DM 27 billion (in 1982 prices, equivalent to about 1.9 percent of GDP). The increase in domestic absorption, however, more than compensates for the lower exports and higher imports, so that GDP rises by about 1.8 percent: owing to the improvement in the terms of trade, real income increases by even more. With capital and land being fixed factors of production, the increase in aggregate output triggers an overproportionately large increase in employment.25

As a result of the deterioration in the external balance, nominal GDP falls by about 1¾ percent. Revenue of the government from direct taxes and social security contributions can be expected to decline by a similar percentage,26 so that the general government deficit increases by a little more than ½ percentage point of GDP.27 With the ratio of investment to GDP up by a little less than ½ percentage point, the share of private savings in GDP declines by about ¾ percentage points. These changes in the savings-investment balance are in line with historical experience. Indeed, the projected increase in the fiscal deficit is of a similar magnitude as the actual increase for 1986–88, and the projected decline in the private savings rate is considerably smaller in absolute amount than the increase that took place in 1985–87.

A better insight into the way the economy adjusts to the exchange rate shock can be obtained from the sectoral results presented in Table 8. The appreciation leads to a large drop in domestic-currency prices of protected goods, the sector with the highest exposure to international trade.28 This sector benefits relatively little from the increase in domestic absorption, and as a consequence it suffers relatively large losses in output and employment. The basic-goods sector is less dependent on foreign demand and is therefore able to maintain output and employment despite the sharp drop in exports and increase in imports. The traded-goods sector, in contrast, benefits from the appreciation despite greater competition from imports because the increase in domestic demand more than compensates for the decline in exports. The clear winner, however, is the nontraded-goods sector. As anticipated, the rise in domestic demand, and the decline in the domestic currency price of those goods that are traded internationally, leads to a significant improvement in the relative price for nontraded goods. This triggers an expansion of output and an increase in employment. Nevertheless, absolute prices fall also in this sector—although by a smaller amount than in the other sectors—because of the lower input and nominal wage costs.

Variable import levies and subsidies in the basic-goods sector neutralize the exchange rate effects on prices in this sector; consequently, prices for basic goods increase sharply relative to those for other goods (and in relation to the baseline simulation) in the presence of these policies (Table 8, scenario 1). This leads to a rise in employment and output in this sector that far exceeds the domestic demand for basic goods. The surplus production has to be dumped in the world market if a buildup of stocks is excluded; hence, exports increase sharply.29 The output and employment gains of the basic-goods sector are at the expense of the other sectors. Higher input and nominal wage costs (because of the assumption of constant real wages and the spillover effects from food prices into the consumer price index) lead to an increase in domestic prices that reduces foreign and domestic demand for products produced in these sectors. The resultant drop in output and employment in these sectors is larger than the rise in the basic-goods sector, so that GDP and aggregate employment decline by about 1.0 percent and 1.6 percent, respectively, in relation to the baseline simulation. Owing to somewhat higher exports and lower imports as well as a larger terms of trade gain, the trade balance declines by about DM 5.5 billion (equivalent to 0.4 percent of GDP) below the baseline simulations. The consumer price level and the real value of the deutsche mark rise by 1.3 percent against the baseline, which implies a corresponding fall in real domestic absorption.

If trade barriers prevent imports in the protected-goods sector from responding to the appreciation, relative prices, output, and employment increase in this sector compared with results in the baseline simulation (Table 8, scenario 2). The effects of these trade restrictions are, however, much smaller than those of the variable import levies and export subsidies used as an instrument of protection in the basic-goods sector. As a consequence, the effects on other sectors and the economy at large are also less severe than under scenario 1. Nevertheless, protectionist practices in the protected-goods sector are likely to result in a decline in GDP and domestic absorption of about 0.2 percent, a fall in aggregate employment of 0.3 percent, and an increase in the real trade balance of DM 2.6 billion (or 0.2 percent of GDP, Table 7) compared with the baseline scenario. In addition to shifting the burden of adjustment to other sectors of the economy, import protection in the protected-goods sector also has the effect of hurting exporting companies in this sector, as evidenced by the fall of almost 1 percent (against the baseline scenario) of exports of these protected goods.

More serious than the protectionist practices analyzed in scenario 2 are the labor market rigidities and regulations, which are assumed to prevent output and employment in the nontraded-goods sector from reacting to the shift in demand toward nontraded goods elicited by the exchange rate change (scenario 3). Indeed, with output failing to respond to higher demand, prices of nontraded goods increase by 5 percent against the baseline, which, because of the assumptions of zero pure profits and equalization of wage increases across sectors, leads to a real wage increase of almost 4.5 percent. Sharply higher wage costs result in higher domestic prices (and a larger real appreciation) than in the baseline simulation and in a deterioration of the international competitiveness of the economy. This leads to a drop in exports, output, and employment. As a result of the lower domestic activity level, imports also fall, and the real trade balance improves by almost DM 5.5 billion (0.4 percent of GDP) relative to the baseline scenario. The effects of the rigidities simulated in this scenario are so strong that real income, GDP, and aggregate employment all fall—not only in relation to the baseline simulation, but also in comparison with the base period.

Finally, when all the protectionist practices and structural rigidities so far considered are combined, the exchange rate shock, which under the baseline simulation led to an improvement in income and employment, has negative implications for the economy. As in the previous scenario, real income, GDP, and employment all decline in relation to both the baseline simulation and the base period. In addition, helped by protectionist trade policies, exports decline by less—and imports by more—than in scenario 3, so that the trade balance improves by almost DM 12 billion (or 0.8 percent) in relation to the baseline simulation. In the baseline case, a 1 percent increase in the real exchange rate led to a DM 4.9 billion deterioration of the real trade balance. In scenario 4, the same real exchange rate change is capable of inducing a change in the real trade balance of only DM 1.6 billion.

IV. The Costs to Adjustment

Although the economy clearly would be better off if more liberal policies were followed, there are some groups of the population who are likely to lose, and it is the strong resistance of those groups that has prolonged protectionist policies and structural rigidities. Table 9 presents the change in the real return to productive factors as a result of the exchange rate shock under the baseline simulation and alternative scenarios. Under scenario 1, the return to capital and land in the basic- goods sector increases substantially over the baseline simulation. Similarly, under scenario 2, the return to capital in the protected-goods sector increases, and under scenario 3 the return to those who are able to hang on to their jobs despite the decline in employment rises sharply. In contrast, when all protectionist practices and structural rigidities are combined, as in scenario 4, it is only the return to capital and land in the basic-goods sector and the return to those who manage to maintain their jobs despite the slowdown of economic activity that are higher than in the baseline simulation. The returns to fixed factors in all other sectors are lower—as is, of course, the return to those who have to rely on unemployment support as a result of the job losses generated under scenario 4. Seen from this perspective, it is not surprising that the strongest resistance against more liberal policies in Germany comes from the representatives of agriculture and the coal-mining industry and from those with a high degree of job security—that is, the employees and workers in the nontraded-goods sector who anticipate an increasing demand for their services.

Table 9.

Real Return to Productive Factors Under the Baseline and Alternative Scenarios

(In percentage changes from base period or baseline)a

article image
Note: Real returns are defined as nominal returns deflated by the consumer price index.

Results for the baseline simulation indicate percentage deviations of endogenous variables from the values they would have attained in the absence of the exchange rate shock. Results for the alternative policy simulations indicate percentage changes from the baseline simulation.

Return to capital and land.

V. Concluding Remarks

The exercise conducted in this paper has illustrated how certain protectionist practices in Germany are capable of impeding the adjustment of the economy to an exogenous exchange rate shock. These distortions are likely to lead to income, output, and employment losses as well as to a slowdown of the adjustment process itself. Inefficient domestic producers are protected at the expense of more efficient foreign and domestic suppliers; in addition to short-term welfare losses, this protection is also likely to have long-term consequences for the growth potential and the dynamics of the economy.

This paper has also illustrated how labor market rigidities and regulations are likely to combine with protectionist polices to generate an environment in which an appreciation of the exchange rate is a serious threat to economic growth and employment. In the absence of these distortions, and with appropriate macroeconomic policies, the economy would react positively to an appreciation of the exchange rate—with output, employment, real income, and absorption all up, and prices and the trade surplus down. Resistance to more liberal economic policies, however, is likely to come from well-organized interest groups that stand to lose if present policies are changed.

REFERENCES

  • Armington, Paul S., “The Geographic Pattern of Trade and the Effects of Price Changes,” Staff Papers, International Monetary Fund (Washington), Vol. 16 (July 1969), pp. 176199.

    • Search Google Scholar
    • Export Citation
  • Burda, M.C., and J.D. Sachs, “Institutional Aspects of High Unemployment in the Federal Republic of Germany,” NBER Working Paper 2241 (Cambridge, Massachusetts: National Bureau of Economic Research, May 1987).

    • Search Google Scholar
    • Export Citation
  • Dixon, Peter B., and others, ORANI—A Multisectoral Model of the Australian Economy (Amsterdam and New York: North-Holland, 1982).

  • Guitián, Manuel, “The Effects of Changes in the Exchange Rate on Output, Prices and the Balance of Payments.” Journal of International Economics (Amsterdam), Vol. 6 (1976), pp. 6574.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lächler, U., “The Elasticity of Substitution between Imported and Domestically Produced Goods in Germany,” Weltwirtschaftliches Archiv (Kiel), Vol. 121 (No. 1, 1985), pp. 7496.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lluch, C., A.A. Powell, and R.A. Williams, Patterns in Household Demand and Saving (New York: Oxford University Press, 1977).

  • Organization for Economic Cooperation and Development (OECD), Costs and Benefits Protection (Paris, 1985).

  • Soltwedel, Rüdiger, and others, Deregulierungspotentiale in der Bundesrepublik (Tubingen: J.C.B. Mohr, 1986).

  • Statistisches Bundesamt, “Input-Output-Tabellen 1982,” in Fachserie 18, Reihe 2 (Stuttgart, Mainz: Kohlhammer, 1987).

  • Weiss, F.D., “Importrestriktionen der Bundesrepublik Deutschland,” Die Weltwirtschaft (Tubingen), No. 1 (1985) pp. 88100.

*

Mr. Mayer, an economist in the Central European Division of the European Department of the Fund, is a graduate of the University of Konstanz and the University of Kiel in the Federal Republic of Germany.

1

Guitián (1976) showed that this is likely to he the case when the share of the nontraded-goods sector in total output is smaller than that of the traded-goods sector.

2

Growth of GNP is given by gnp=(DD/GNP)dd+(X/GNP)x(M/GNP)m, with DD denoting domestic demand, X exports, M imports, and lowercase letters standing for percentage changes. Partial analysis would suggest that GNP falls in response to an appreciation because of the negative influence from trade. In contrast, general equilibrium analysis suggests that there are forces that militate against a decline of GNP. For instance, the improvement in the terms of trade and the decline in the price level elicited by the appreciation are likely to boost real private consumption through income, wealth, and real balance effects. Higher real consumer expenditures and the need to improve competitiveness in the traded-goods sector would support real investment, and a fiscal policy that follows medium-term nominal expenditure targets would temporarily lead to higher real public consumption.

3

There are several impediments to labor mobility, both on the demand and on the supply side. On the demand side, laws enacted for the protection of employees have over time increased the costs of layoffs for companies, with the result that employers have become reluctant to hire even when their business outlook is improving; often they prefer to invest with a view to increasing the productivity of the existing work force to meet higher demand. On the supply side, relatively generous unemployment benefits and a deeply rooted regionalism have tended to discourage laid-off workers from bearing the adjustment costs associated with a change to a new industry, region, or both. See Burda and Sachs (1987).

4

Industries with similar characteristics were aggregated into sectors to facilitate the quantitative analysis that follows in Sections III and IV.

5

Among the service industries listed in Table 2, only telecommunications and transportation were allocated to the traded-goods sector, given the potential (telecommunications) or actual (transportation) role of these industries in the international trade in services.

6

The telecommunications and transportation industries that were classified as traded-goods industries accounted for about 12 percent of the value added of this sector.

7

In the manner of Armington (1969), it is assumed that domestically produced goods and imported goods are imperfect substitutes.

8

Note that only sector 1 (basic goods) uses land as a productive factor.

9

The German input-output table has a “government sector” that creates the goods used by government. This sector is contained in the nontraded-goods sector of the model.

10

The Armington (1969) assumption implies that goods are differentiated by source of production. Thus, German exporters face a downward-sloping foreign demand curve.

11

Implicit in this treatment of import supply is the assumption that the rest of the world produces goods for the German market that are not complete substitutes for goods shipped to other markets. In other words, the Armington (1969) assumption also holds for import supply.

12

The assumption of zero pure profits implies that there are no undistributed profits. Rents to productive factors, however, are not excluded.

13

Note that market clearing does not necessarily imply full employment of all factors. In fact, as outlined below, the model allows the fixing of factor prices at above full employment levels and the equilibrating of supply and demand at these prices.

14

In the following experiments it is assumed that the shares of (private and public) consumption and investment in domestic absorption remain unchanged.

15

Using the relationship between per capita GDP and w estimated by Lluch, Powell, and Williams (1977, p. 248).

16

The design of the model ensures that there is a unique solution provided that the matrix to be solved is not singular (see Dixon and others (1982)).

17

Real wages that are rigid and above full employment levels have been a common characteristic of most European countries during the 1980s.

18

In principle, the decline in the domestic price level elicited by the exchange rate appreciation will increase real financial household balances, which in turn will boost real consumption and contribute to a decline in the savings rate. Higher real consumer demand, together with a higher rate of return for capital, will then stimulate real investment. The role of monetary policy in this context is to support these effects by stabilizing nominal private absorption through monetary expansion in line with potential output growth and some acceptable and sustainable rate of price increases.

19

Although these assumptions are in accord with the key principles of the European Community Common Agricultural Policy and the German support scheme for coal mining, the implementation of these policies is in reality slightly different. For example, the authorities are likely to take into account, at least partially, the effects of an appreciation on costs and factor incomes in the respective sectors when they set prices there. Moreover, subsidies are given to two main industrial users of coal—the steel and electricity industry—to compensate them partially for the use of high-priced German coal. In the case of steel, the subsidies are borne by the federal budget, and in the case of electricity the costs of the subsidies are passed on to the consumers (Kohlepfennig). These arrangements may alleviate the direct effects of the coal support scheme on the respective industries, but they do not change the final effects on the German economy. In either case, relative prices are distorted, and aggregate production costs increase.

20

Protection for shipbuilding, which has been afforded largely in the form of subsidies, was not separately modeled. Instead, it was assumed that measures similar to those in the other protected industries were taken to reduce competition from imports.

21

This assumption, which again is only a crude approximation of reality, reflects the low degree of intersectoral wage differentiation (see Table 3) and the highly centralized organization of the German trade unions. This organization has contributed to the fast diffusion of wage increases from one industry to the other industries in the economy.

22

Note that the increase in real domestic absorption is not explicitly explained by the model. The result is, however, in line with conventional wisdom: real private consumption increases as a result of higher real income and real balance effects, elicited by the decline in the price level and the expansionary effects of a monetary policy that follows nominal potential GDP growth rather than actual growth: real investment increases because of accelerator effects and a higher return to capital; and real government expenditures increase as the authorities stick to their medium-term nominal expenditure targets.

23

The buoyancy of imports with respect to GNP is slightly greater than 2.0 and close to the historical average in Germany over the recent years. The (general equilibrium) elasticity of imports with respect to the real exchange rate is about 0.7 percent, a little higher than most econometric estimates of the partial real exchange rate elasticity of imports.

24

The elasticity of exports with respect to the real exchange rate is about 1.0. Econometric studies have usually arrived at a value a little smaller than this. Part of the discrepancy can be explained by the assumed absence of protectionist practices and structural rigidities under the baseline simulation, which contributes to higher trade elasticities in the economy. Indeed, in scenario 4, the export and import elasticities decline to 0.9 and 0.2, respectively.

25

Given the assumption of constant returns to scale, the percentage change in output equals the weighted average of the percentage changes of factor inputs plus any efficiency gains that may arise from a reallocation of factors among industries. In the case of Germany, there are only very small differences in the marginal returns to labor across industries, so that efficiency gains do not arise. Thus, the increase in employment is almost identical to the increase in output divided by the share of labor in total value added (0.63).

26

Indirect taxes will remain largely unaffected because they are related to nominal domestic absorption, which remains unchanged.

27

This implies a largely neutral fiscal policy stance, since the decline in total government revenue is smaller than that of nominal GDP and the ratio of expenditures to potential nominal GDP remains broadly unchanged.

28

Like the basic-goods sector, the protected-goods sector has a higher price elasticity of imports and exports, and a higher substitution elasticity between imported and domestically produced goods, than does the traded-goods sector. The export orientation of the protected-goods sector (measured as exports in percent of output) is only a little smaller than that of the traded-goods sector, but more than twice as high as that of the basic-goods sector. Also, its import penetration ratio is only slightly smaller than that of the basic-goods sector but substantially larger than that of the traded-goods sector.

29

In reality, part of the increase in production would, of course, lead to stock accumulation and therefore to a smaller rise in exports than projected in this illustrative exercise.