Macroeconomic policies and adjustment in Yugoslavia : some counterfactual simulations

WDP-16 1 6 z World Bank Discussion Papers Macroeconomic Policies and Adjustment in Yugoslavia Some Counterfactual Simulations Fahrettin Yagci Steven Kamin iFILE CtO PY ' 1   16 U World Bank Discussion Papers Macroeconomic Policies and Adjustment in Yugoslavia Some Counterfactual Simulations Fahrettin Yagci Steven Kamin The World Bank Washington, D.C.  The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printingJuly 1987 Discussion Papers are not formal publications of the World Bank. They present preliminary and unpolished results of country analysis or research that is circulated to encourage discussion and comment; citation and the use of such a paper should take account of its provisional character. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. Because of the informality and to present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The most recent World Bank publications are described in the catalog N\Tew Publications, a new edition of which is issued in the spring and fall of each year. The complete backlist of publications is shown in the annual Index of Publications, which contains an alphabetical title list and indexes of subjects, authors, and countries and regions; it is of value principally to libraries and institutional purchasers. The latest edition of each of these is available free of charge from Publications Sales Unit, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'1ena, 75116 Paris, France. Fahrettin Yagci, an economist in the Economic Analysis and Projections Department of the World Bank when the paper was written, is now in the Programs Department of the South Asia Regional Office. Steven Kamin, a consultant to the Bank, is at the Massachusetts Institute of Technology. Library of Congress Cataloging-in-Publication Data Yagci, Fahrettin, 1944- Macroeconomic policies and adjustment in Yugoslavia. (World Bank discussion papers ; 16) 1. Yugoslavia--Economic policy--1945- 2. Yugoslavia--Economic conditions--1945- 3. Balance of payments--Yugoslavia. I. Kamin, Steve. II. Title. III. Series. HC407.Y6Y34 1987 339.5'09497 87-16064 ISBN 0-8213-0938-2  ABSTRACT After a period of successful growth in the 1970s, Yugoslavia was hit very hard by a severe balance of payments crisis at the end of the 1970s that has overshadowed the economic scene since. What were the causes of the crisis? Could it have been averted through the adoption of alternative policies in the 1970s? What can be said about the adjustment policies pur- sued since 1980 in the light of the lessons of the 1970s? These questions are addressed in this paper with the help of some counterfactual policy simulations for the period 1973-79, using a macroeconomic model developed for the Yugoslav economy. The simulations yielded three main conclusions. First, the payments crisis was attributable to a deterioration in the policy environment, par- ticularly with respect to the exchange rate, wages, and interest rates, although external shocks and/or excessive expansion of domestic demand have played an aggravating role. Second, the crisis could have been avoided with alternative macroeconomic policies. Third, Yugoslavia could have achieved external balance and satisfactory growth in investment and output since 1980 had the policy environment been reformed promptly and significantly.  ACKNOWLEDGEMENTS We would like to thank Wafic Grais, Mansour Farsad and Laura Tyson for many valauble discussions, EPDCO staff for useful comments, Barbara Ossowicka for the help with the data and computer work, and Vicky Sugui for skillful typing. The paper expresses our views alone and should not be interpreted as reflecting the opinions of the World Bank, the Massachusetts Institute of Technology or the above individuals.  TABLE OF CONTENTS Page No. 1. INTRODUCTION . ................. ...... . 1 2. FROM GROWTH TO CRISIS................ 2 3. POLICY WEAKNESSES AND DISTORTIONS . . 7 4. THE MODEL ................. .... ......10 5. ALTERNATIVE POLICIES FOR YUGOSLAVIA IN THE 1970's . .20 6. CONCLUSIONS ...................... 27   1. INTRODUCTION After a period of successful growth in the 1970s, Yugoslavia was hit very hard by a severe balance of payments crisis at the end of the 1970s that has overshadowed the economic scene since. What were the causes of the crisis? Could it have been averted through the adoption of alternative policies in the 1970s? What can be said about the adjustment policies pursued since 1980 in the light of the lessons of the 1970s? These questions are addressed in this paper with the help of some counterfactual policy simulations for the period 1973-79, using a macroeconomic model developed for the Yugoslav economy. The simulations yielded three main conclusions. First, the payments crisis was attributable to a deterioration in the policy environment, particularly with respect to the exchange rate, wages, and interest rates, although external shocks and/or excessive expansion of domestic demand have played an aggravating role. Second, the crisis could have been avoided with alternative macroeconomic policies. Third, Yugoslavia could have achieved external balance and satisfactory growth in investment and output since 1980 had the policy environment been reformed promptly and significantly. Traditionally, excessive expansion of demand has been regarded as a prime culprit in foreign exchange problems. Inappropriate expenditure-switching policies have also been pointed to, although their importance has been considered to be only secondary. As a consequence, stabilization programs have emphasized the importance of reducing expenditures in responding to balance of payments difficulties. Under such programs, however, excessive reductions in expenditures with insufficient expenditure-switching and structural adjustment might  -2- result in a substantial sacrifice of employment and growth that would lead to social unrest and abandonment of the program. By focusing specifically on some expenditure-switching policies, this paper also addresses this issue. In the context of Yugoslavia's experience, it analyzes the role of the exchange rate policy in keeping the balance of payments in check without sacrificing employment and growth. The rest of the paper is structured as follows. Section 2 provides a brief description of the economic policies and growth of the 1970s. The core of the model is sketched in Section 3, while Section 4 evaluates alternative growth paths the economy would have taken under different exchange rate, wage, and interest rate policies than those pursued in 1973-79. Section 5 summarizes the main findings of the paper. 2. FROM GROWTH TO CRISIS Yugoslavia chose to maintain a policy of high growth after the 1973 oil price increase and during the consequent turbulence in the world economy. The rate of growth of gross social product (GSP) 1/ amounted to 6.1 percent, on average, in the 1973-79 period, slightly higher than that of the preceding seven years (Table 1). This growth was driven by an acceleration of capital expenditures: fixed investment rose to a rate of 8.0 percent in 1973-79 from the 6.0 percent 1/ GSP is the counterpart of CDP in the Yugoslav accounting system. It yields somewhat lower figures than GDP does as it excludes part of services such as education, health, public administration, and bank- ing and insurance.  - 3 - Table 1: YUGOSLAVIA -- MAIN MACROECONOMIC DATA 1966-72 1973-79 1980-85 Growth Rates in Real Terms (Z) GSP 5.9 6.1 0.9 Domestic Absorption 5.7 6.4 -3.1 Gross Fixed Investment 6.0 8.0 -7.8 Total Consumption 5.7 5.6 -0.9 Employment a/ 2.0 4.2 2.5 Labor Productivity a/ 3.8 1.8 -0.2 Net Income per Worker a/ 5.1 3.4 -6.7 Exports of Goods 7.0 2.9 4.0 Imports of Goods 10.8 7.5 -7.5 Implicit GSP Deflator 11.2 17.8 45.2 Money Supply (M1) 4.9 10.0 -8.0 Share in GSP (X) Total Investment 38.1 42.2 43.1 Gross Fixed Investment 30.5 34.1 27.5 Changes in Stocks 7.6 8.1 15.6 Domestic Savings 37.0 39.8 43.1 Current Account Balance -1.1 -2.4 -0.5 Others Real Interest Rate on Saving Deposits (Z) n.a. -9.6 -18.8 Maximum Real Lending Rate (%) n.a. -7.8 -10.8 Source: Federal Statistical Office, Statistical Yearbook, various issues, and National Bank of Yugoslavia, Quarterly Bulletin, various issues. a/ In the social sector.  of 1966-72. In the same period, the ratio of total investment to GSP increased from 30.5 percent to 34.1 percent. Yugoslavia's ambitious investment drive, which was aimed at moderating the unemployment problem, was not, however, matched by a corresponding increase in savings. The rate of domestic savings rose from 37.0 percent in 1966-72 to only 39.8 percent in 1973-79. The rising level of investment therefore had to be financed by foreign savings. The composition of aggregate demand changed substantially in the 1970s. The share of exports of goods and non-factor services in aggregate demand fell from 24.2 percent in 1973 to 17.7 percent in 1979, while the share of imports of goods and non-factor services remained constant at 29.0 percent. As a result, the current account deficit in CSP rose from 1.1 percent in 1966-72 to 2.4 percent in 1973-79, reaching a peak of 6.0 percent in 1979. The increased liquidity in the international capital markets made borrowing a convenient substitute for improving the policy environment. However, by the end of the 1970s, the country was facing a severe balance of payments crisis, total debt having increased from $4.3 billion in 1973 to $13.7 billion in 1979. The structure of the terms of the debt also changed substantially: the average maturity and grace period fell from 21.4 and 4.2 years to 14.9 and 3.5 years, respectively, in the same period. Creditworthiness eroded, and external financing dried up. Since 1980, although severe austerity measures have been introduced, the macroeconomic policies have not been reformed  significantly. Consequently, the country has bogged down in a long period of stagflation. What were the causes of the payments crisis? This question has been raised repeatedly by interested parties within and outside Yugoslavia, since a correct identification of the main reasons is imperative to designing an effective program of adjustment to restore stability and growth. Contrary to the general perception, the deterioration in the terms of trade induced by the first oil shock does not appear to have been a significant contributor to the crisis. The terms of trade worsened by an annual average rate of only 0.5 percent in 1973-79, which is substantially lower than the annual average deterioration of 1.2 percent in all non-oil developing countries in the same period. Moreover, as Table 2 indicates, after falling 13.0 percentage points in 1974, the terms of trade gradually improved. This moderate deterioration may be explained by the fact that the share of oil in total imports was only 3.9 percent in 1973. In addition, the bulk of Yugoslavia's exports were comprised of manufactured products for which prices were rising. The short-lived world recession in 1974/75 was also not a main suspect, given that the build-up of the crisis occurred over a longer period and continued after the recession ended. These considerations indicate that the payments difficulties that emerged in 1979 were attributable more to policy-induced domestic shocks than to exogenous external ones. On the domestic front, the acceleration of fixed investment obviously contributed to the  Table 2: YUGOSLAVIA -- SOME ECONOMIC INDICATORS (percent unless otherwise indicated) Real Growth Ratio of Index of Index of Interest Real Rate Current Terms Index of Labor Real Income Rate on Maximum of Fixed Account of Real Exch. Product- per Savings Lending Invest- Balance Trade Rate a/ ivity Worker Inflation Deposits Rate ment to GSP 1972 100.0 100.0 100.0 100.0 15.2 -8.9 -7.2 3.1 2.8 1973 102.0 84.8 102.6 98.9 18.9 -10.7 -9.9 2.9 2.5 1974 89.0 70.4 106.2 109.4 22.5 -14.3 -13.5 9.1 -4.6 1975 93.0 72.2 104.4 108.4 19.2 -11.0 -9.2 9.7 -3.4 1976 94.0 72.1 105.2 113.6 13.4 -5.2 -3.4 8.1 0.7 1977 93.0 66.8 108.3 115.7 14.7 -6.7 -4.7 9.4 -3.6 1978 96.0 63.7 110.7 124.1 14.9 -6.8 -3.9 10.5 -2.3 1979 95.0 58.4 113.7 126.2 20.7 -12.6 -9.7 6.4 -6.0 1980 94.0 63.5 112.6 114.6 30.3 -22.1 -18.3 -5.9 -4.0 1981 93.0 70.4 111.0 109.3 40.2 -32.0 -28.2 -9.8 -0.9 1982 97.0 81.4 109.0 105.2 31.8 -22.3 -9.8 -5.5 -0.7 1983 95.0 111.0 105.5 93.6 40.8 -21.7 -3.5 -9.7 0.4 1984 96.0 124.1 105.8 86.8 52.6 -10.1 -1.0 -8.2 1.1 1985 97.0 127.2 104.9 83.3 74.4 -4.4 -4.4 -9.5 1.2 Source: Federal Statistical Office, Statistical Yearbook, various issues, National Bank of Yugoslavia, Quarterly Bulletin, various issues, and IMF, International Financial Statistics, various issues. a/ Nominal exchange rate multiplied by the ratio of GNP deflator of USA to GSP deflator of Yugoslavia.  - 7 - deterioration of external payments. A less ambitious investment drive would have resulted in more favorable balance of payments, although it would also have generated lower growth and employment. Besides, as the data indicate, the destabilizing element was not the level of demand but its composition: the rate of net exports to aggregate demand rose from -4.8 percent in 1973 to -11.3 percent in 1979. Behind this significant switch away from exportables was the substantial worsening of the policy environment in general and the rapid appreciation of the exchange rate in particular. Indeed, one of the main points coming out of this paper is that the payments crisis of 1979 may be ascribed to the accumulated effects of the misconceived exchange rate, wage, and interest rate policies. 3. POLICY WEAKNESSES AND DISTORTIONS The nominal exchange rate was almost constant from 1972 to 1979, whereas in the same period domestic prices tripled. Consequently, the real exchange rate appreciated continuously, averaging 7.7 percent a year in 1973-79, and the index of the real exchange rate fell from 100.0 in 1972 to 58.4 in 1979 (Table 2). This dramatic loss of competitiveness contributed substantially to the fall in the growth rate of exports from 7.0 percent in 1966-72 to 2.9 percent in 1973-79 (Table 1) . Distortions in the factor markets are observed in both the labor and the capital markets. The average rate of growth of labor productivity in the social sector was 1.8 percent in 1973-79. However, real wages in the same sector grew at an average rate of 3.4 percent in that period, an indication of a substantial margin between productivity  - 8 - and wage payments. On the cost side, wage payments in excess of productivity exerted pressure on prices, a situation that led to lower competitiveness. On the demand side, higher wages shifted the distribution of income in favor of social sector wage earners and possibly affected the rate of domestic savings adversely. Real interest rates on both deposits and loans were grossly negative (Table 2). The low rates for deposits discouraged financial savings and the repatriation of workers' remittances. The low rates for lending gave rise to excess demand for credit and necessitated credit rationing, which interfered with its efficient allocation. Other policy weaknesses of a more structural nature compounded the effects of the deteriorating macroeconomic environment in the 1970s. Although discussed here briefly, these weaknesses are not included in the counterfactual simulations because their impact on the performance of the economy is harder to quantify. The effective rate of protection is an appropriate indicator for measuring the trade-related distortions in the product market. There is evidence that the variation in effective protection among subsectors was high. The average tariff rate (tariff and all tariff- like changes) was around 15.0 percent in 1973-79. However, the rates varied between 0.0 and 35.0 percent; in general, those for intermediates were substantially lower than the rates for final products. In addition, the use of quantitative restrictions was widespread. All these factors point to higher effective rates and significant variation across products. Until recently, most prices have been determined administratively in Yugoslavia by Self-Management Agreements among the  - 9 - parties involved. This rigid price system, coupled with the high and differential levels of protection, created substantial distortions in the product market and affected efficiency adversely. Solidarity among the socially owned firms led to widespread socialization of financial losses and enabled firms to operate under very soft budget constraints. 1/ Because of the lack of financial discipline, economic agents were not held accountable for their actions, and non-viable enterprises could continue to operate. In addition, investment decisions were not guided by appropriate criteria for project selection. Consequently, the flow of resources into the most profitable activities was seriously constrained. Factor mobility among the regions was very limited, including capital, labor and foreign exchange. The reason is that firms in Yugoslavia are regionally based, and they tend to establish vertically integrated production units within their own regions to obtain the maximum possible benefits from intraregional contacts (securing credit, getting licenses, etc.). The resultant regionalism has hindered the dissemination of technical and managerial skills, created excess capacity, constrained specialization and the division of labor, produced large interregional differences in productivity and income, and severely affected overall efficiency and productivity. 1/ Under these circumstances the elasticity of response to changes in incentives is likely to be low.  - 10 - 4. THE MODEL The model abstracts from sectoral details and input-output relationships in order to focus sharply on interactions among the major macroeconomic performance indicators -- output, trade and current account balances, prices, savings, and investment -- and the major macro policy instruments -- exchange rate, wages, and interest rates. Three essential elements dominate the structure of the model. First, only one type of good is assumed to be produced domestically, and it is either exported abroad or marketed internally, depending on the relative prices of exports and domestically sold output. Consumers, firms, and the government have a demand for both this good and for final imports, depending on income and relative prices. In the "flex-price" closure of the model, national output and the prices of domestically marketed goods are determined so as to satisfy simultaneously the country's foreign exchange constraint and the condition of zero excess demand in the market for domestically marketed output. In the "fix-price" formulation, either the foreign exchange constraint or the zero excess demand condition is relaxed, depending on the assumption made regarding Yugoslavia's access to foreign credit. Second, output is assumed to depend strongly on the use of imported intermediate goods. In cases where foreign capital inflows are not accommodative, this condition gives the balance of trade and payments exceptionally important roles in determining output. Third, the model incorporates a number of institutional features of the Yugoslav economy in the 1973-79 period that depart strongly from conventional assumptions about price, wage, and interest rate flexibility. In Yugoslavia, employment and wages are both, in  - 11 - large part, institutionally determined social objectives, and so both are modeled as exogenous policy instruments. Interest rates and the allocation of credit to firms are determined administratively as well, and so they are also modeled as exogenous policy instruments rather than as the outcome of supply and demand in the financial markets. Finally, it was decided that the practice of price determination in the Yugoslav self-managed enterprise system in the 1970s is best approached through a mark-up model rather than through a framework of competitive market pricing. 1/ The model's major components are summarized and interpreted below. Production. The economy produces only one good, denoted Q, using labor (L), capital (K), and an imported intermediate good (N). We assume the production function to be of the Leontief fixed-proportion type, separable into each of the three inputs: Q = Q(L,K,N) = min [1(L), k(K), n(N)]. (1) Because of the growing problem of unemployment in Yugoslavia, it is assumed that the labor constraint is never binding. It is also assumed that the capital constraint is potentially binding, so that at all times Q < QP = k(K), (2) 1/ With the liberalization of some prices in the last two years, price behavior has changed substantially. The suggested mark-up model is obviously not appropriate to explain the current inflation.  - 12 - where QP stands for potential output. However, once installed, the capital stock is fixed, so that production may at times plausibly fail to reach the capital constraint. Finally, because imported intermediates are both scarce and variable, the relationship between output and imported intermediates is considered to be binding at all times. Thus, Q = nN. (3) Real income, Y, is defined as Y = Q - N. (4) Supply Behavior. Given total output, the share that suppliers export abroad depends on the ratio of the domestic price, P, to the currency-converted, subsidy-adjusted export price, PE = ER*(l + su)*PEW, where ER represents the exchange rate (in dinars per dollar), su the subsidy rate, and PEW the exogenous world price of the good. Accordingly, real exports, E, are determined as E = E(PE/P,Y) SE/6(PE/P) > 0, 6E/SY > 0. (5) Similarly, the supply of domestically marketed output is determined as SQ = Y - E = Y - E(PE/P,Y) = SQ(PE/P,Y) (6) SSQ/6(PE/P) < 0, 6SQ/6Y > 0.  - 13 - Demand Behavior. World demand for Yugoslavia's good is considered to be perfectly elastic at the world price, PEW. As equation (3) indicates, domestic demand for imported intermediates, N, is responsive only to output, not to relative prices. Domestic demand for the imported final goods, FM, and domestically produced and sold goods, DQ, depend on both their relative prices and total absorption, or expenditure A: FM = FM(PM/P,A) 6FM/6(PM/P) < 0, SFM/6A > 0 (7) DQ = A - FM = A - FM(PM/P,A) = DQ(PM/P,A) (8) 6DQ/6(PM/P) > 0, 6DQ/6A > 0, where PM = PMW*(l + tf)*ER represents the domestic currency price of imported final goods, tf the tariff rate, and PMW the exogenous world price of the final imports. Total absorption in the model depends on the endogenous consumption of households, HC, as well as on a number of exogenous expenditure components -- government consumption, CC, government investment, GI, and economic sector investment, EI: A = HC + EI + GC + GI. (9) Household consumption, HC, in turn depends (through a number of distributional equations) on income, Y, as well as on the real interest rate r: HC = HC (Y,r) 6HC/6Y > 0, SHC/6r < 0. (10)  - 14 - The real interest rate is defined as r = - P, (ll) where P represents the rate of growth of (actual) prices and i is a weighted average of nominal interest rates for different deposit categories (a dot on a variable indicates a percentage change). Total absorption, imports, and the demand for domestic goods can now be summarized: A = A[Y,r(i,P),EI, GC, GI] (12) FM = FM[PM/P, A(Y,r(i,P), EI, GC, GI)] (13) DQ= DQ[PM/P, A(Y,r(i,P), EI, GC, GI)]. (14) Price Behavior. Under the Yugoslav system of self-management, the price determination of the 1970s appears to be approached best through an administered-price model, with firms charging a mark-up over prime costs in setting their prices. In this model, domestic prices, P, are determined as a multiplicative mark-up over wages, intermediate input costs, and working capital costs, inclusive of indirect taxes: P = (1 + mr) (1 + t) [W*L/Q + r I*KW/Q + PN*N/Q], (15) where mr : Mark-up rate t : Indirect tax rate  - 15 - W : Nominal wage rate L : Employment r1 : Lending rate Kw : Working capital N : Imported intermediate input PN : Domestic currency price of imported intermediate input With mr and t kept constant, this relationship is converted into the estimated equation in growth rates actually used in the model: P = sw*W + sk*r; + sn*PN, (16) where the estimated coefficients, sw, sk, and sn are interpreted as the shares of labor, working capital, and imported intermediates in prime cost, respectively. Key Equilibrium Relationships. The core of the model consists of three relationships between output and relative prices. First, the demand for imports, N + FM, along with net services payments, NFI, must equal the flow supply of foreign exchange provided by exports, E, reserve drawdowns, R, and net capital inflows, CI: E + CI + R = N + FM + NFI (17) or E(P/PE,Y) + CI + R = N(Y) + FM[P/PM,A(Y,r,EI,GC,GI)] + NFI. (18)  - 16 - For given CI, R, and NFI, increases in income, Y (and hence output), raise both exports and imports. Our data confirm that the marginal propensity to import out of income exceeds the positive response of exports to increases in output, so that imports expand more than exports in the event of a rise in Y. Accordingly, if the foreign exchange constraint is to hold when income grows, relative prices, P/PM and P/PE, must fall to discourage imports and encourage exports. With RP representing either P/PE or P/PM, since both PE and PM are independent of Y, the locis of (RP,Y) combinations that represent points of external balance are depicted as curve EB in Figure 1. The second key relationship in the model equates the demand for domestically marketed output, DQ, with the supply of domestically marketed output, SQ: SQ(PE/P,Y) = DQ[PM/P,A(Y,r,EI,CC,GI)]. (19) Increases in income increase both the supply and demand for domestically marketed output. Our data indicate that the marginal propensity to purchase domestically produced goods out of income is exceeded by the marginal propensity to supply them. This pattern corresponds to a stable system in a basic Keynesian framework. Accordingly, increases in income are associated with the development of excess supplies of domestically marketed goods at the same relative price; the price of domestic goods relative to final imports must therefore drop to raise domestic demand, lower domestic supplies, and re-establish internal balance. In Figure 1, IB represents the locis of (RP,Y) points at which internal balance holds, all else being equal.  - 17 - The third key relationship in the model is the price determination equation already described above. Since none of the determinants of the domestic price, P, and hence the relative price, RP, is related to the level of income or output, the RP curve depicting this condition in Figure 1 is a horizontal line. Model Solution under Different Closures. Obviously, all three of the relationships described in Figure 1 cannot be binding at once. Depending on the type of closure deemed appropriate, either the external balance condition, EB, the internal balance condition, IB, or the price relationship, RP, must be relaxed. In a setting with flexible prices and instantaneously clearing markets, movements in production costs, demand for goods, and output would be reflected in real income and relative prices rather than in the aggregate level of domestic prices. In that case, the RP relationship would be eliminated, so that equilibrium would occur at point A in Figure 1, the point at which both the foreign exchange constraint is met and the supply of domestically marketed output equals its demand. As the flexible price scenario appears to be especially inappropriate to Yugoslavia's economy in the 1970s, the RP relationship is retained, and either the internal balance or external balance condition is residualized. The choice of closure for the model is influenced most by Yugoslavia's access to foreign credit. In the 1970s, Yugoslavia had relatively easy access to foreign credit, and combined extensive stimulation of demand with substantial borrowing overseas. As the capital account was essentially accommodative of the pressures on the current account, the external balance condition did not bind. This  - 18 - Figure 1 RP .~~~~ F) RPI 11?, *1' Figure 2 oP.P : \ \ + ~~~A- t - R P, NN~~~~ 1B 1B .~~~~~~~ Y%  - 19 - closure is achieved in our model by endogenizing the level of net capital infLows, CI. Equilibrium under this closure occurs at point B in Figure 1, with the fixed level of relative prices, RP, determining output along the IB curve. Note that at point B, the relative price is above the EB curve; this result means that the level of exports are lower, and the level of imports and the trade deficit higher than the levels that would be consistent with the foreign exchange constraint as shown for some given amount of net capital inflows. After 1979, Yugoslavia's prior accumulation of foreign debt, coupled with the generally less favorable environment for international commercial lending, made further external borrowing untenable. Yugoslavia's foreign exchange constraint thus became binding, so that the appropriate model closure provides for an exogenously specified level of net capital inflows; Yugoslavia was forced to take corresponding steps to restrain the level of domestic absorption. In our model, the internal balance, IB, constraint is relaxed by residualizing fixed investment, El, in total absorption. A = HC(Y,r) + EI + CC + CI. (20) EI is predetermined when the EB constraint is relaxed, but is residualized under the binding EB closure, so that the IB constraint is then relaxed. Equilibrium occurs at point C in Figure 1 in this closure.  - 20 - 5. ALTERNATIVE POLICIES FOR YUGOSLAVIA IN THE 1970s The framework elaborated above vividly highlights the impact of Yugoslavia's macroeconomic policies on its current account balance and foreign debt over the course of the 1970s. Figure 2 depicts the Yugoslav economy as it was initially in the decade -- at a point such as A in 1972 -- on the IB curve on account of the accommodating capital inflows. High investment, low interest rate, and high wage growth policies stimulated total expenditures and shifting the IB curve to the right and the EB curve downward. At the same time, high wage growth, caused the real exchange rate to appreciate over time. Hence, the Yugoslav economy moved from A in 1972 to a point such as B in 1979 that is considerably farther removed from the locis of external balance. Theoretically, the Yugoslavian authorities could have offset the undesirable effects of their investment program with sufficiently stringent exchange rate, wage, and interest rate policies. It is not clear a priori whether "reasonable", "non-extreme" actions in these macro policy areas could have substantially moderated the growth of Yugoslavia's foreign debt. This section describes a variety of counterfactual simulations intended to assess whether reasonable, alternative macro policies could have enabled Yugoslavia to pursue its investment program of the 1970s without incurring the external financial burden it experienced at the onset of the 1980s. The model is solved, in every case, under the open balance of payments closure, a choice that reflects Yugoslavia's easy access to foreign credit during the period. The levels of both private and government investment are kept at their historical levels for the period, as are employment, all world prices, workers' remittances, and reserve movements. Nominal exchange rate,  - 21 - wage, and interest rate policies are the instruments that are changed in the alternative counterfactual scenarios described below. 1/ Scenario 1 (Si): Constant Real Exchange Rate. The real exchange rate (measured in terms of the exchange rate-adjusted domestic price level and the US GNP deflator) appreciated at a rate of 7.7 percent a year in the 1973-79 period. By holding the real exchange rate constant at its 1973/74 value and keeping all other exogenous variables unchanged, this scenario simulates the effects of an active exchange rate policy. The nominal exchange rate is allowed to adjust to keep the real exchange rate at the specified level. Scenario 2 (S2): Higher Real Interest Rates. Ex post real interest rates were highly negative in the 1973-79 period. S2 simulates the effects of higher rates for both loans and deposits. The nominal rates are gradually increased until the ex post real rates become positive in 1979. Scenario 3 (S3): Lower Wage Growth. Real wages in Yugoslavia grew faster than productivity in 1973-79. S3 is designed to simulate the effects of a nominal wage rate set to equate the growth of the real wage to that of labor productivity over the period. A summary of the simulation results is provided in Table 3. The model has been calibrated to equate its base case solution for the period 1973-79 with actual historical values for those years. 2/ 1/ The coefficients of the behavioral equations are also kept constant across the scenarios. It might be argued that each scenario represents a different policy regime and that the coefficients (and the values of some of the exogenous variables) might have changed under each scenario. This argument might be correct, but there is no way to estimate the hypothetical coefficients that would have prevailed under each policy regime. 2/ Most of the behavioral equations in the model are estimated econometrically.  Table 3: SUMMARY OF SIMUtATION RESULTS Actual S1 S2 S3 S4 Average Average Averdge Average Average 1973-79 1979 1973-79 1979 1973-79 1979 19/3-79 1979 1973-79 1979 Growth Rate (a) GSP 6.1 7.0 6.5 7.2 5.5 3.9 6.2 7.2 6.2 6.4 Household Consumption 5.5 5.8 5.1 5.7 4.4 0.8 5.0 5.7 4.6 4.0 Exports 2.9 4.5 5.4 5.1 1.7 -1.4 4.1 5.7 5.3 4.6 Imports 7.5 18.1 6.9 17.6 7.5 17.0 7.1 17.6 6.7 16.8 Infldtion 17.7 20.7 18.5 20.9 18.8 26.3 16.1 19.2 22.4 40.8 Ratios (M) Current Account/GSP -2.4 -6.0 0.1 -1.1 -2.6 -7.0 -0.1 -3.6 0.2 -1.0 Debt Service Ratio 22.2 27.2 19.8 20.6 23.2 30.2 20.5 23.4 19.8 20.7 Household Income/GSP 63.7 62.2 60.3 59.1 61.4 60.9 62.0 60.6 60.2 58.9 Household Saving Rate 14.5 14.1 14.7 14.2 14.9 14.9 14.7 14.3 14.9 14.8 Domestic Saving Rdte 35.4 37.2 36.7 40.0 35.7 38.2 36.3 38.7 36.9 40.5 Others (S billion) Gross Capital Intlow 2.6 3.9 1.4 0.8 2.7 4.6 1.9 2.2 1.4 0.7 Total Externdl Debt 8.0 13.8 5.6 6.0 8.5 15.2 6.4 9.1 5.6 5.8  - 23 - SI: Constant Real Exchange Rate. The simulation results for S1 indicate that a constant real exchange rate policy could have substantially moderated the growth of Yugoslavia's foreign debt over the course of the 1970s. The ratio of the current account to GSP registers an average 0.1 percent for 1973-79 in the S1 simulation, as compared to Yugoslavia's actual experience of -2.4 percent. As a result, the simulated total foreign debt figure for Yugoslavia in 1979 is $6.0 billion, as compared to the actual $13.8 billion, while the simulated debt service ratio in that year is 20.6 percent as compared to the actual figure of 27.2 percent. Under S1, both expenditure-switching and expenditure-reduction effects cause Yugoslavia's external imbalance to decline. Expenditure- switching results from the increases in the prices for exports and imports (measured in domestic currency) relative to domestic prices, a trend that encourages exports and discourages imports at every level of output and absorption. In 1973-79, Yugoslavia's exports and imports actually increased at an annual average of 2.9 percent and 7.5 percent, respectively. Under the constant real exchange rate scenario, the rates of growth of real exports and imports are 5.4 percent and 6.9 percent, respectively. These figures indicate substantial expenditure-switching. The improved current account performance in S1 also reflects reductions in domestic expenditures. Real private consumption grows at a 5.1 percent rate between 1973 and 1979, compared with Yugoslavia's actual performance of 5.5 percent. The decline in real consumption growth, in turn, reflects the impact of heightened inflation on real wage growth over this period. By increasing the domestic currency cost of imported inputs, the constant real exchange rate policy raises  - 24 - Yugoslavia's average rate of inflation in 1973-79 from 17.7 percent (actual) to 18.5 percent (Si). Because the path of nominal wages is fixed in Si, the growth of real wages falls. As may be seen in Table 3, this effect lowers the share of household income in GSP from 63.7 percent (actual) to 60.3 percent (Si) over the period and therefore helps increase the rate of domestic savings from 35.4 percent to 36.7 percent. The shifts in the distribution of income that accompany the rise in inflation more than offset the effects of the lower real interest rates in depressing the savings rate. In the face of the drop in domestic demand, the improved performance of output is explained by substantially greater net exports. S2: Higher Real Interest Rates. A higher path of real interest rates (for both deposits and loans) over the 1970s could have benefitted the Yugoslav economy in two ways. First, it might have increased the rate of savings and hence improved the current account. Second, higher interest rates might have acted to screen out less productive investment projects and hence have improved the efficiency of investments; this effect is not incorporated in the Yugoslavia model, which is more narrowly macroeconomic in focus, but it should be kept in mind. Notwithstanding the theoretical desirability of higher interest rates, the simulation results for S2 indicate that a policy of a high interest rate would not only have produced stagflation, but would actually have worsened the current account and debt accumulation over the 1970s. These effects result largely from the impact of higher lending rates on the prime costs of production and hence domestic prices. The rate of inflation for 1979 rises from an actual level of  - 25 - 20.7 percent to 26.3 percent in S2, compared with 20.9 percent in Scenario 1. While the average rates of inflation for the 1973-79 period are comparable for S2 (18.8 percent) and Si (18.5 percent), the average real exchange rate appreciates considerably more in S2, since the path of the nominal exchange rate is kept fixed at historic levels. As a result, export growth falters significantly under S2; this situation reinforces the impact of lower real wages (caused by the higher inflation) on consumer spending, and the growth in total output falls to an average of 5.5 percent for the period, compared to the actual 6.1 percent. Increases in household savings resulting from higher real deposit rates also contribute to the fall-off in expenditures. The cumulative effects of the loss of competitiveness are considerable. Average export growth for the period falls to 1.7 percent, while import growth amounts to 7.5 percent. As a consequence, the current account balance rises to an average of -2.6 percent over the period, and total debt increases to $15.2 billion in 1979, which is $1.4 billion more than the actual debt. Under the current specification of the model, the only favorable effect that higher interest rates have is a slightly greater rate of domestic savings than that realized in 1973-79. The improved savings are attributable to both increased deposit rates and income redistribution toward enterprises. Two points are worth mentioning. First, S2 confirms that if efficiency gains are not considered and if prices are determined mainly by costs and a constant mark-up, then an isolated active interest rate policy might be stagflationary. Second, as will be seen in the discussion of S4, a compensatory devaluation to neutralize the  - 26 - inflationary effects of the higher interest rates could eliminate the undesirable trade effects of the financial reforms. S3: Lower Wage Growth. In this scenario, the growth of nominal wages is set so as to equate the growth of real wages with the growth of the endogenously determined output/labor ratio. In response to the slower nominal wage growth under S3, inflation averages only 16.1 percent for the period, compared to the actual 17.7 percent. The lower rates of inflation in turn slow down the over valuation of the real exchange rate and hence demonstrably improve Yugoslavia's external balance. In S3, Yugoslavia's current account balance averages only -0.1 percent of GSP, compared to the historical average of -2.4 percent for 1973-79, while accumulated foreign debt is $9.1 billion in 1979 rather than the actual $13.8 billion. These gains in external balance in Scenario 3 are not achieved without cost. Because of the direct impact of the policy of low wage growth on household income, the growth of real private consumption drops to 5.0 percent annually for 1973-79, lower than both Yugoslavia's actual 5.5 percent growth and Si's 5.1 percent. The resultant decrease in total domestic expenditures just offsets the increase in activity originating in the export sector, so that, output growth is left about unchanged from its actual level. With output unchanged and the balance of trade improved, the rate of domestic savings also improves under S3. S4: A Combined Scenario (S4). S4 represents a policy package: a combination of Si, 2, and 3. Under S4, the response of the external sector to the inflationary impact of the policy of high interest rates (S2) and the deflationary impact of the policy of low wage growth (S3) are essentially nullified by the policy of an active real exchange rate  - 27 - (SI). Both S2 and S3 have largely contractionary effects on demand and output, but they are more than offset by the impact of the real exchange rate in boosting exports and discouraging imports. Real private consumption grows at a rate of only 4.6 percent in 1973-79 in S4, compared to Yugoslavia's historical average of 5.5 percent. The contribution of S2 and S3 to the outcome of the policy package is evident in the improved domestic saving rate in particular: the rate of domestic savings under S4 is 40.5 percent in 1979, compared to the actual 37.2 percent. This improvement is the result of both to the higher rates for deposits and the redistribution of income toward enterprises. The current account responds well to both the expenditure- switching and expenditure-reduction policies: it moves from the actual deficit of 2.4 percent to a simulated surplus of 0.2 percent over the period. The simulated level of foreign debt in 1979 amounts to $5.8 billion, compared to the actual level of $13.8 billion. As a consequence, the debt service ratio in 1979 falls from the actual rate of 27.2 percent to 20.7 percent. 6. CONCLUSIONS The results of our simulation experiments strongly indicate that the explosive growth of Yugoslavia's foreign debt in the 1970s and the consequent severe balance of payments crisis were not inevitable. Nor would Yugoslavia have been forced to sacrifice its investment program and high growth and employment targets in order to achieve external balance. If Yugoslavia had avoided the deterioration in its policy environment then despite unfavorable external conditions, its  - 28 - foreign indebtedness at the end of the decade would have been less than half the actual value. The alternative policies, moreover, are not "extreme", "lunreasonable", or "restrictive" in any sense. Stable real exchange rates, positive real interest rates, and the equating of real wage growth with productivity growth are conditions associated with any normal, well-functioning economy. Note also that the favorable response of the economy to the alternative policies simulated with the model are not the result of using unrealistically high elasticities (the econometrically estimated elasticities used are more on the conservative side, with the elasticity of export supply and import demand with respect to relative prices very close to unity). Rather, it is attributable to the correction of some of the policy mistakes made in macroeconomic management. A detailed evaluation of the recent economic policies would fall outside the scope of their paper. However, the simulation results provide some basis for assessing the main adjustment strategy Yugoslavia has adopted since 1980. Failure to take corrective actions promptly forced Yugoslavia to rely exclusively on reductions in expenditures to control the payments crisis. Since 1980, fixed investment and real income in the social sector have declined by 7.8 percent and 6.7 percent a year, respectively. The growth of output has stagnated around 0.9 percent as a reaction to the protracted contraction in demand (Table 1). Policy reforms have been considerably delayed. Compared to its 1972 value, the exchange rate was overvalued in the first three years of adjustment (Table 2), real interest rates are still negative, and the intended reforms related to financial discipline, investment criteria,  - 29 - and foreign exchange and credit allocation are not yet in place, six years after the outbreak of the crisis. The simulation experiments for 1973-79 indicate that external balance could had been achieved without sacrificing investment and growth if the right policies had been in place. The inference is that Yugoslavia could have combined external balance with satisfactory growth of investment and output if the policy environment had been reformed promptly and significantly after 1979. Finally, our results confirm the usefulness of simulation models and scenario analysis in evaluating the interactive effects of separate policy instruments in combined policy packages. Note, for example, that, taken in isolation, the higher real interest rate path tested in Scenario 2 produced some unfavorable results. On the other hand, when combined with an active real exchange rate policy, its deleterious impact on real exchange rates was muted, and it was freed to raise savings rates and, to a certain extent, to lower the share of household income.  DISTRIBUTORS OF WORLD BANK PUBLICATIONS ARGENTINA FRANCE KENYA SPAIN Carlos Hirsch, SRL World Bank Publcations Africa Book Service (E.A.) Ltd. Mundi-Prensa Libros, S.A. Galeria Guemes 66 Avenue d'Iena PI 0. Box 45245 Castello 37 Florida 165, 4th Floor-Ofc. 453/465 75116 Paris Nairobi 28001 Madrid 1333 Buenos Aires GERMANY, FEDERAL KOREA, REPUBLIC OF SRI LANKA AND AUSTRALIA, PAPUA NEW REPUBLIC OF Pan Korea Book Corporation THE MALDIVES GUINEA, FIJI, SOLOMON UNO-Verlag P. 0. Box 101, Kwangwhamun Lake House Bookshop ISLANDS, AND VANUATU D-5300 Bonn I Seoul P.O. Box 244 Info-Line Simrockstrasse 23 KUWAIT 100, Sir Chittampalam A. Gardiner Overseas Document Delivery GREECE MEMRB Mawatha Box 506, GPO KEME P.O. Box 5465 Colombo 2 Sydney, NSW 2001 24, Ippodamou Street MALAYSIA SWEDEN AUSTRIA Athens-11635 ~ ~ ~ ~ ALASIAFor single titles: AUSTRIA Athens-11635 Unuversity of Malaya Cooperative ABCE Fritzes Kingl. Gerold and Co. GUATEMALA Bookshop, Limited Hovbokhandel GA111Wen 3 Librerias Piedra Santa P.0 o 17 aa atiBm Regerinigsgatan 12, Box 16356 Graben 31 Centro Cultural Piedra Santa Kuala Lumpur S-103 27 Stockholm BAHRAIN II calle 6-50 zona I MEXICO For Subscription orders: MEMRB Information Services Guatemala City INFOTEC Wennergren-Wiiiiams AB P.O. Box 2750 HONG KONG, MACAU San Fernando No. 37 Box 30004 Manama Town 317 Asia 2000 Ltd. Col. Torielo Guerre S-104 25 Stockholm BANGLADESH 6 Fl., 146 Prince Edward Road, W, Thalpan, Mexico D.F. SWITZERLAND Micro Industnes Development Kowloon MOROCCO Librairie Pavot Assistance Society (MIDAS) Hong Kong Societe d'Etudes Marketing 6 Rue Grenus G.P.O. Box 800 HUNGARY Marocaine Case postal 381 Dhaka Kultura 2 Rue Mobere, Bd. d'Anfa CH 1211 Geneva 11 BELGIUM P.O. Box 139 Casablanca TANZANIA Publications des Nations Unies 1389 Budapest 62 THE NETHERLANDS Oxford University Press Av. du Roi 202 INDIA Medical Books Europe, BV (MBE) P. 0. Box 5299 1060 Brussels For single titles Noorderwal 38, Dar es Salaam BRAZIL UBS Publishers' Distributors Ltd. 7241 BL Lochem THAILAND Publbcacoes Tecncas Post Box 7015 NEW ZEALAND Central Department Store Internacionais Ltda. New Delhi 110002 Hills Library and Information 306 Siom Road Rua Peixoto Gomide. 209 10 First Main Road Service Bangkok 01409 Sao Paulo, SP Gandhi Nagar Private Bag CANADA Bangalore 560009 New Market TRINIDAD & TOBAGO, Le Diffuseur Apeejav Chambers, P.O. Box 736 Auckland ANTIGUA, BARBUDA, C.P. 85, 1501 Ampere Street 5 Wallace Street NIGERIA GRENADA, GDMYANA, boucherville, Quebec bomba, 400001 University Press Limited MONTSERRAT, ST. KITTS J4B 5E6 81-B, Chownnghee Lane Three Crowns Building Jericho AND NEVIS, ST. LUCIA, CHILE Calcutta 700016 Private Mail Bag 5095 ST. VINCENT & GRENADINES Editorial Renacimiento 7/188, I(A). Swarup Nagar Ibadan Svstematics Studies Unit Miratlores 354 Kanpur 208001 NORWAY 55 Eastern Main Road Santiago Sivaganga Road Tanum Karl Johan, A.S. Curepe COLOMBIA Nungambakkam P. 0. Box 1177 Sentrum Trinidad, West Indies Madras 600034 Oslo I TUNISIA Carrera 6 No. 51-21-A Rajendra Nagar Societe Tunisienne de Diffusion Bogota D.E. Patna 00016 Mirza ook Agenc 5 Avenue de Carthage Apartado Aereo 4430 For subscripton orders 65, Shahrah-e-Quaid-e-Azam Tunis Cali, ValUe Universal Subscription Agency P.O. Box No. 729 TURKEY Pvt. Ltd. Lhr ae iaeiAS COSTA RICA 18-19 Community Centre Saket Lahore 3 Ha9e itapevi as. Libreria Trejos New Delhi 110 017 PERU 469. Istiklal Caddesi Cavre 11-13 INDONESIA Editonal Desarrollo SA Bevoglu-Istanbul San JosePt. Indira Limnited Apartado 3824 UGANDA Jose ~~~~~~~~JI. Sam Ratulangi 37 Lima Uganda Bookshop COTE D'IVOIRE Jakarta Pusat THE PHILIPPINES P.O Box 7145 Centre d'Edition et de Diffusion P.O. Box 181 National Book Store Kampala Africaines (CEDA) IRELAND 701 Rizal Avenue UNITED ARAB EMIRATES 04 B.P. 541MerMaiaMMBGlCo Abidjan 04 Plateau TDC PublishersMerMaiaM RBGlCo 12 North Frederick Street PORTUGAL P. 0. Box 6097 CYPRUS Dublin I Livrana Portugal Sharah MEMRB Information Services Rua Do Carmo 70-74 UNITED KINGDOM P.O. Box 2098 ISRAEL 1200 Lisbon Microinfo Ltd. Nicosia The Jerusalem Post The Jerusalem Post Building SAUDI ARABIA, QATAR P 0. Box 3 DENMARK P.O. Box 81 Janr Book Store Alton, Hampshie GU 34 2PG SamfundsLitteratur Romena Jerusalem 91000 P. 0. Box 3196 England Rosenoemns Alle 11 ITALY Riyadh 11471 VENEZUELA DK-1970 Fredeniksberg C. Licosa Commissionaria SINGAPORE, TAIWAN, Libreria del Este DOMINICAN REPUBLIC Sansoni SPA BURMA, BRUNEI Aptdo. 60.337 Editora Taller, C. por A. Via Lamarmora 45 Information Publications aracas Restauracion Casella Postale 552 Private, Ltd. WESTERN SAMOA Apdo. postal 2190 50121 Florence 02-06 Ist Fl., Pei-Fu Industrial Weslev Bookshop Santo Domingo Bldg., 24 New Industnal Road P.O. Box 207 JAPAN Singapore Apia EGYPT, ARAB REPUBLIC OF Eastern Book Service Al Ahram 37-3, Hongo 3-Chome, SOUTH AFRICA YUGOSLAVIA Al GaJaa Street Bunkvo-ku 113 For single titles Jugoslovenska Knjiga Cairo Tokyo Oxford Universitv Press YU-11000 Belgrade Trg Republike FINLAND JORDAN Southern Afrca ZIMBABWE Akateeminen Kirjakauppa Jordan Center for Marketing Cape Town 8000 Textbook Sales Pvt. Ltd. P.O. Box 128 Research Box 3799 SF-00101 P.O. Box 3143 For subscription orders Harare Helsinki 10 Jabal International Subscription Service Hrr Amman P.O. Box 41095 CraighaU Johannesburg 2024   It The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'I6na Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome (5 Telephone: (202) 477-1234 Telephone: (1) 47.23.54.21 Chiyoda-ku, Tokyo 100,Japan Telex: WUI 64145 WORLDBANK Telex: 842-620628 Telephone: (03) 214-5001 RCA 248423 WORLDBK Telex: 781-26838 Cable Address: INTBAFRAD WAS HINGTON DC -t C) C) CD CD n ISSN025-21O ISN 0-21309382 DOO1