Yugoslavia - Adjustment policies and development perspectives

WORLD PUB-3954 A WORLDNAK COUNTRY STUDY YUGOSLAVIA Adjustment Policies and Development Perspectives FILE COPY   A WORLD BANK COUNTRY STUDY YUGOSLAVIA Adjustment Policies and Development Perspectives The World Bank Washington, D.C., U.S.A.  Copyright 0 1983 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. First printing October 1983 All rights reserved Manufactured in the United States of America World Bank Country Studies are reports originally prepared for intemal use as part of the continuing analysis by the Bank of the economic and related conditions of its developing member countries and of its dialogues with the governments. Some of the reports are published informally with the least possible delay for the use of governments and the academic, business and financial, and development commu- nities. Thus, the typescript has not been prepared in accordance with the proce- dures appropriate to formal printed texts, and the World Bank accepts no responsi- bility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. Any maps used have been prepared solely for the convenience of the readers; the denominations used and the boundaries shown do not imply, on the part of the World Bank and its affiliates, any judgment on the legal status of any territory or any endorsement or acceptance of such boundaries. The full range of World Bank publications is described in the Catalog of World Bank Publications; the continuing research program of the Bank is outlined in World Bank Research Program: Abstracts of Current Studies. Both booklets are updated annu- ally; the most recent edition of each is available without charge from World Bank Publications in either Washington or Paris (see the back cover for addresses). Library of Congress Cataloging in Publication Data Main entry under title: Yugoslavia, adjustment policies and development perspectives. (A World Bank country study) Includes bibliographical references. 1. Yugoslavia--Economic policy--1945- 2. Yugoslavia--Economic conditions--1945- I. World Bank. II. Series. HC407.Y82 1983 338.9497 83-6551 ISBN 0-8213-0189-6  - iii - YUGOSLAVIA ADJUSTMENT POLICIES AND DEVELOPMENT PERSPECTIVES This report is based on the findings of a World Bank economic mission which visited Yugoslavia in June 1981. The mission consisted of the following; Suman Bery, Chief of mission Petros Aklilu (Agriculture) Deepak Bhattasali (Foreign Trade and Finance) Kosara Gavrilovic (Translator and Interpreter) Paul Harrison (Consultant) (Agriculture) Frederick Kilby (Employment and Regional Development) Francois Laporte (Industry) Surinder Malik (Industry) Leslie Manison (IMF) (Foreign Trade and Finance) Mieko Nishimizu (Total Factor Productivity; Regional Development) Sherman Robinson (Adjustment Policies; CGE Model) Laura Tyson (Consultant) (Adjustment Policies; CGE Model) Assistance to the mission's work in Washington was provided by Jeffrey Lewis and Mukaila Ojelade. Ann Pepper had primary responsibility for document production. A draft of the report was discused with official and academic bodies in Yugoslavia in June and July 1982, and additional material has been included based upon that visit. The report was finalized and distributed to the World Bank's Executive Directors in November 1982. The facts and judgments provided here are current as of that date.  - iv - Currency Equivalents 1/ 1976 1 US Dollar = 18.19 Dinars 1 Dinar 0.055 US Dollars 1977 1 US Dollar 18.30 Dinars 1 Dinar 0.055 US Dollar 1978 1 US Dollar 5 18.64 Dinars 1 Dinar 0.054 US Dollar 1979 1 US Dollar = 19.00 Dinars 1 Dinar 0.053 US Dollar 1980 1 US Dollar = 24.91 Dinars 1 Dinar 0.040 US Dollar 1981 1 US Dollar 35.51 Dinars 1 Dinar 0.028 US Dollar 1/ Period average exchange rates. The dinar has not been maintained within announced margins since July 12, 1973. Following a devaluation on October 22, 1982, the rate stood at 63.51 Dinars per US Dollar. Glossary of Abbreviations BOAL Basic Organization of Associated Labor COAL Complex (or Composite) Organization of Associated Labor CGE Computable General Equilibrium CIFER Community of Interest for Foreign Economic Relations COI Community of Interest DOD Debt Outstanding and Disbursed DRC Domestic Resource Cost GMP Gross Material Product ICOR Incremental Capital-Output Ratio KBP Kosovska Banka Pristina LDR Less Developed Region(s) MDR More Developed Region(s) MLT Medium- and Long-Term OAL Organization of Associated Labor QR Quantitative Restriction SSE Small Scale Enterprise(s) YBIEC Yugoslav Bank for International Economic Cooperation Throughout this report, the term 'region' is used to refer to Yugoslavia's six republics and two autonomous provinces.  -v - YUGOSLAVIA ADJUSTMENT POLICIES AND DEVELOPMENT PERSPECTIVES Table of Contents THE MAIN REPORT Page No. COUNTRY DATA INTRODUCTION PART I: ADJUSTMENT PERFORMANCE AND POLICIES I. The Adjustment Strategy of the 1976-80 Plan .......... 1 A. Background to the Plan ..................... 1 Economic Trends . * * ........ . .... . ............. . 1 Institutional Change ....................... 3 B. The Adjustment Strategy ................. ......... 6 II. Plan Performance: Overview and External Sector Developments ...*... 13 A. An Overview ...................................... 13 B. The External Sector .......**.*.a.........*...... 15 Sources of Change in the Current Account Deficit ..................... ....... 15 Merchandise Exports ... ................ . . . . . . . . . . . . . . 19 Performance in Developed Country Markets ....... 22 Invisibles .............. ......... ....................... 24 External Debt .................................. 28 III. Plan Performance: The Domestic Economy ........... ... 35 A. Sectoral Growth Performance ............ * ..... .... 35 B. Investment ............... ............... 39 C. Capital Efficiency ......... . .. ............. .. .......... 42 D. Saving Performance ...... ......................... 47 Aggregate Trends .... ......................... .... 47 Household Savings ......... ........... .. .... 48 Enterprise Savings ..* . ........... 49 IV. The 1976-80 Plan: The Policy Framework .............*. 55 A. Investment Allocation ........ .................... 55 B. The Foreign Trade and Payments Regime ............ 66 Institutional Developments ................... .. . 66  - vi. - Table of Contents (continued) Page No. Incentive Framework ...... ..................... 67 Foreign Exchange Allocation ............. ........ 72 Export Incentives ............. .. ............... 75 Trade Bias .............................. 79 C. Exchange Rate Policy ...... ................ 84 D. Summary Assessment ...... ......................... 93 V. The 1981-85 Adjustment Strategy ......... ............ 97 A. Overview ......... ............................... 97 B. Medium-Term Prospects . ........................... 107 Sectoral Implications ................... 112 Capital Account Issues ............ ............. 115 Policy Issues ....... ........................ 117 PART II: DEVELOPMENT PERSPECTIVES AND ISSUES VI. The Industrial Sector ............................. 120 A. Introduction ............................... 120 B. The 1976-80 Plan ..... .............. 122 Aggregate Targets ................... 122 Investment ........ .. ......... 123 Output ................... 124 Employment .................... .... 125 Imports ........ .. ......... 126 Exports ..................................... ... 127 C. The 1981-85 Plan .......... ........ 128 Priority Industries . 134 D. Export Potential, Markets and the Incentive System 144 Plan Targets ... 144 Past Performance ........... 145 Export Markets .... 146 Export Incentive System . 149 Export Promotion Infrastructure . 152 E. Some Policy Issues in Industry . 152 Incentive Policy ........ 153 Export Credit .. 156 Investment Choice .. 156  - vii - Table of Contents (Continued) Page No. VII. Agriculture ................ ........................... 159 A. Structure of the Agriculture Sector ...... ........ 159 Resource Base .. ................................ 159 Land Tenure and Production Orientation ......... 159 Agricultural Population .. ..................... 160 Role of Agriculture in the Economy ............. 160 B. The Agriculture Sector 1976-1980: Plan and Performance ..................................... 162 Plan Objectives and Strategy .................. . 162 Assessment of Performance ......... ............ 163 Policy Framework ............... ................ 163 Agricultural Pricing and Subsidy Policies ...... 163 Investment Allocation ............ .............. 165 Association of Farmers ............. ............ 167 Land Reclamation and Abandonment ....... ........ 168 Agricultural Exports ...... ..................... 169 C. The 1981-85 Plan ...........* *................. 170 Evaluation of Plan Targets and Measures ........ 170 Land Use ..... . ........ ...................... 171 Food Processing Capacities ..................... 176 Summary Evaluation ....... ...................... 177 VIII. Employment ................ 180 A. Introduction .. . ............ ... 180 B. Labor Force Trends 1971-75 .............. . .... . 180 C. The 1976-80 Plan ................................. 184 Social Sector Employment ............. .......... 184 Private Nonagricultural Employment ............. 194 Private Sector Agricultural Employment ......... 195 Unemployment ................................... 203 Overall Performance 1976-80 ............ . ....... 208 D. Labor Force Balances 1970-80 ............ . ...... 210 E. Regional Employment Issues ............ ........... 212 Regional Demographic Trends .......... .......... 215 External Migration ............................. 218 Interregional Migration ... ..................... 220 F. The 1981-85 Plan ..*.*.*..***.....******..... *.... 225 IX. Regional Development Issues ...... .................... 233 Introduction ......................................... 233 A. Regional Trends in Output, Population and Per Capita Incomes ...... ........................ 234  - viii - Table of Contents (Continued) Page No. B. The Framework of Regional Policy .......... ....... 242 The Scale and Impact of Regional Transfers ... o. 242 Other Regional Policy Measures ..... ............ 245 C. Productivity Growth and Development Policies in the LDR ......................................... 246 The Sources of LDR Growth 1965-78 .... .......... 248 Changes in the Efficiency of Production ........ 253 The Role of Enterprises in Productivity Growth . 261 Regional Planning ... ............................. 262 Competition Policy ............................. 264 The Impact of Regional Fiscal Policy on Resource Allocation .................................... 267 D. Recent Regional Policy Initiatives and Suggestions for Further Policy Reform ..... .................. 271 Domestic Joint Ventures ..... ................... 271 Policy Issues ..................... 273 An Action Program for Kosovo . .................. 278 X. Adjustment Policies and Development Perspectives .... 283 A. Stabilization and Trade Policies ............... o 284 B. Investment and Financial Policies .............. o 285 C. Institutional Reform ...... ..................... 289 D. Employment and Regional Development Policies .... 290 E. Concluding Remarks ............... ............... 290 METHODOLOGICAL AND STATISTICAL ANNEX 293 LIST OF TEXT TABLES Table No. Chapter I 1.1 External Sector Developments, 1965-80 ........... 4 1.2 External Sector Targets of the 1976-80 Plan ..... 6 1.3 Structure of Investment in Fixed Assets in Economic Sectors 1971-75 and 1976-80 ........ 9 1.4 Macroeconomic Indicators 1971-75 (Actual) and 1976-80 (Planned) ...... ....................... 10 1.5 Estimated External Financing Requirements of the 1976-80 Plan .... .............................. 11  -ix - List of Text Tables (Continued) Table No. Page No. Chapter II 2.1 Main Indicators of the 1976-80 Period ........... 14 2.2 Sources of Change in Current Account Deficit, 1975-80 ................. . ............... 16 2.3 Evolution of the Real Deficit, 1975-80 .......... 18 2.4 Aggregate Regional and Commodity Distribution of Exports, 1970, 1975 and 1979 .................. 20 2.5 Growth of Yugoslav Exports by Market and Commodity Group, 1970-79 ........... ........... 21 2.6 Constant Market Share Analysis of Yugoslav Exports to Developed Country Markets, 1970-79 . 24 2.7 Developments in the Invisibles Account, 1976-80 . 25 2.8 Migration and Workers' Remittances .. ............ 26 2.9 Shipment and Transportation Services ............ 27 2.10 Tourism Services .. * ... ................... 28 2.11 External Debt Indicators, 1975-80 ............... 29 2.12 Structure and Terms of Borrowing, 1975-80 ....... 30 2.13 Five Year Time Profile Ratio of Medium- and Long-Term External Debt (Disbursed and Outstanding) ... .. ....................... 32 Chapter III 3.1 Actual and Target Growth Rates by Sector, 1971-80 ...... ...................... 36 3.2 Growth of Raw Material and Processing Subsectors in Priority Industrial Sectors .... ............ 37 3.3 Ratio of Direct Imports of Intermediate Inputs to Gross Domestic Output Production ........... 38 3.4 Indicators of the Overall Investment Effort, 1971-80 ..... 41 3.5 Functional Distribution of Real Investment Expenditures, 1971-80 .. ..................... 42 3.6 Incremental Capital-Output Ratios, 1971-80 ...... 44 3.7 Indicators of Aggregate Saving Performance, 1971-79 ........ ............................... 48 3.8 Household Saving Rates (1971-80) .. .............. 49 3.9 Saving Behavior of Social Sector Enterprise, 1971-80 ........................................ 50 3.10 Distribution of Value-Added of Social Sector Enterprises in the Productive Sector .......... 52 3.11 Indicators of Social Sector Enterprise Saving Behavior . ............................ .. 53  -x- List of Text Tables (Continued) Table No. Page No. Chapter IV 4.1 Price Trends in Yugoslav Industry, 1970-79 ...... 56 4.2 Indicators of Relative Market Profitability Across Sectors, 1976-80 57 4.3 A Comparison of Market and Economic Profitability Rates Across Sectors in 1980 .................. 60 4.4 Imports by Import Regime, 1971-80 ................ 69 4.5 Ratio of Imports to Supply Available for Domestic Use, 1972-79 ......................... 70 4.6 Indicators of Import Elasticity, 1971-1980 ...... 71 4.7 Average Quantity Rationing and Foreign Exchange Premium Rates, 1976-80 .74 4.8 Rental Income From Foreign Exchange Rationing as a Percentage of Sectoral Value Added, 1976-80 76 4.9 Estimated Average Export Subsidy Rate, 1976-80 79 4.10 Estimated Sectoral Domestic Resource Costs, 1980 ..81 4.11 Exchange Rates and Indices of External Competitiveness. .... 86 4.12 Actual and Market Clearing Exchange Rates, 1976-80... 89 4.13 Alternative Estimates of Market Clearing Exchange Rates, 1980 .......................... 91 Chapter V 5.1 Major Targets of the 1981-85 Federal Plan ....... 106 5.2 Projected Growth of Macroeconomic Aggregates, 1981-85 ......109 5.3 Foreign Exchange Rationing Under Alternative Scenarios. 110 5.4 Historical and Projected Sources of Changes in Demand for Manufacturing Output, Selected Periods .................0... 112 5.5 Sources of Change in Demand for Manufactured Output, Selected Countries and Periods .114 5.6 Projected Growth Rates of Output in the Manufacturing Sector, 1981-85 .115 5.7 Capital Account Projections, 1981-85 ........... 116  - xi - List of Text Tables (Continued) Table No. Page No. Chapter VI 6.1 Structure and Growth of Industrial Production 121 6.2 Industrial Sector--Growth and Structure Southern European Selected Countries ......... 121 6.3 Industrial Sector Performance, 1976-1980 ....... 122 6.4 Structure and Growth of Industrial Investment in Fixed Assets ....... ....................... 123 6.5 Development of Basic Raw Materials .... ......... 124 6.6 Industrial Employment ............. .............. 125 6.7 Structure and Growth of Industrial Imports, 1975-80 * .................................. 126 6.8 Structure and Growth of Industrial Exports, 1975-80 .......... ............................ 127 6.9 Plan Targets in Industry ..... .................. 129 6.10 Plan Output Targets in the Industrial Sector ... 129 6.11 Planned Investments in Priority Sectors ........ 131 6.12 Plan Projections of External Trade ............. 133 6.13 Supply and Demand Projections for Finished Steel Products .135 6.14 Nonferrous Metals: Existing and Planned Capacities ........ ........................... 138 6.15 Plan Targets in the Chemicals Sector .... ....... 140 6.16 Plan Projections for Main Chemical Product Groups ............ .... 141 6.17 Plan Targets for Metalworking Industries ....... 143 Chapter VII 7.1 Percentage Share of Agriculture and Industry in Social Product ....... ...................... 161 7.2 Agricultural Performance, 1971-1980 .... ........ 162 7.3 Fertilizer Consumption - Active Substance, 1976-1980 .......... ........................... 165 7.4 Livestock Investment Efficiency ..... ........... 166 7.5 Planned and Actual Investment in Agriculture and Agroindustries 1976-1980 .166 7.6 Growth Rates for Major Agricultural Products, 1981-85 ................. ...171 7.7 British Standard Labor Requirements for Yugoslav Private Agriculture, 1980 ..... ................ 175 7.8 Trade Targets of 1981-85 Plan ..... ............. 175  - Ki - List of Text Tables (Continued) Table No. Page No. Chapter VIII 8.1 Employment and Labor Force, 1970-75 ............ 182 8.2 Social Sector Employment 1971-75, and Plan Targets 1976-80 ..... .... s........ 185 8.3 Growth of Social Sector Output and Employment by Sector 1971-75, 1976-80 188 8.4 Rates of Growth of Employment, Captial Stock and Capital Intensity of Production, 1971-79 190 8.5 Growth of Labor Productivity and Real Personal Incomes Per Worker 1971-80 192 8.6 Employment Trends in Private Sector Agriculture 1971-80 .. 197 8.7 Distribution of Households by Economic Activity and Income Per Active Family Member 1973 and 1978 ......... 202 8.8 Some Indicators of Unemployment 1975-80 ........ 204 8.9 Structure of Registered Job Seekers 1970-80 .... 206 8.10 Employment and Labor Force, 1975-80 ............ 209 8.11 Labor Force Flows 1970-80 ........... ........#.... 211 8.12 Share of Social Sector Employment in Total Labor Force by Region 1975 and 1980 213 8.13 Selected Regional Employment Indicators, 1975 and 198CI ............................................ 214 8.14 Regional Demographic Trends, 1961-81 ........... 216 8.15 Absorption of the Natural Increase in the Labor Force, 1976-80 ......... ........ ....... 217 8.16 External Migrants by Region of Origin, 1971 and 1981 ...........*s.* ... .**.... ... 219 8.17 Incomes of Social Sector Employees by Region, Selected Years, 1965-79 223 8.18 Anticipated Changes in the Structure of the Labor Force, 1981-85 ...................... 225 8.19 Plan Targets for Social Sector Output, Employment, Labor Productivity and Personal Incomes, 1981-85 ...................... 238 8.20 Regional Employment Growth Targets 1981-85 230  - siii - List of Te4t Tables (continued) Table No. Page No. Chapter IX 9.1 Regional Trends in Output, Population and Per Capita Output 1952-65, 1965-68 235 9.2 Regional Disparities in Income Per Capita, 1954-80. . 237 9.3 Household Income Per Capita Differentials, 1978 238 9.4 Demographic Indicators, 1950-78 .240 9.5 Scale of Interregional Transfers, 1980 243 9.6 Contribution of Federal Fund to Investment in Less Developed Regions, 1976-80 .245 9.7 Some Basic Indicators of Regional Social Sector Productivity in 1978 .............*......... 247 9.8 Sources of Social Sector Growth for Major Sectors, by Less Developed Region, 1965-78 251 9.9 Percentage Distribution of Social Sector Value- Added According to Relative Levels of Technical Efficiency by Region (1965-1978) 254 9.10 Financial Results of LDR Social Sector Enterprises, 1980 ........$ .................... 264 9.11 Regional Variations in Tax Policy, 1979 ......... 269 9.12 Regional Labor Costs in the Social Sector, 1980 277 List of Charts 1 Bosnia-Herzegovina Potential and Actual Social Sector Social Product (Value-Added), 1965-68 257 2 Kosovo Potential and Actual Social Sector Social Product (Value-Added), 1965-78 ......258 3 Macedonia Potential and Actual Social Sector Social Product (Value-Added), 1965-68 259 4 Montenegro Potential and Actual Social Sector Social Product (Value-Added), 1965-68 .260   - xiv - COUWfTY DATA - YUGOSLAVIA AREA POPULATION DENSITY !3,804 sq. km. 22.3 illion (-id-1980) 86 pereons per sq. km. Rate of Grovthl 0.9X (from 1970 to 1980) 154 persons per eq. km. of agricultural land POPULATION CHARACTERISTICS (1980) HEALTH (1979) Crude birth Rate (per 1,000) 16.9 Population per physician 1,795 Crude Death Rate (per 1,000) 8.6 Population per hospital bed 167 Infant Mortality (per 1,000 live births) 32.6 INCOME DISTRIBUTION (1978) DISTRIBUTION OF LAND OWNERSHIP (1971) S ere oi housieholdincome, lowest quintile 6.6 I owned by top 102 of owners highest quintile 38.7 (social aector Kombinats) 15.1 S owned by smallest 1OX of owners (privata esmllholdera) 84.9 ACCESS TO PIPED WATER (1978) ACCESS TO ELECTRICITY Dwellings with piped water (S) 40.5 i of all dwellings (1978) 89.0 rural (1971) 80.0 NUTRITION (1977) EDUCATION Per capita Calorie Supply 3,445 Adult Literacy rate (2) 85 (1975) (136I of requirement) Primary school enrollment (S) 100 (1977) Per cepita protein supply (grams/day) 101 Secondary school enrollment (2) 79 (1977) GNP PER CAPITA IN 1981 1/% US$2790 GROSS DOMESTIC PRODUCT IN 1981 AVERAGE ANNUAL RATES OF GROWTH (S, constant prices) US $ Mln. S 1970-75 1975-80 1981 GDP at Market Prices 69,617 100.0 6.5 5.8 2.2 Total Consumption 45,651 65.6 6.9 5.4 -1.0 Gross Domestic Investment 25,427 36.5 5.5 5.4 1.3 Gross National Savings 24,677 35.4 6.2 5.0 -12.2 Exports of Goods and Nonfactor 6.7 5.3 10.9 Services 16,354 23.5 6.7 3.2 -3.0 Imports of Goods and Nonfactor Services 17,816 25.6 OUTPUT, LABOR FORCE AND PRODUCTIVITY IN 1980 GDP at Current Factor Cost (1980) Labor Force 4/ Value Added Per Active Resident Worker US Min. X Mlnn S USS (1980) XIa Agriculture 2/ 8,516 13.7 2.50 30.6 3,261 43.2 industry 3/ 28,801 46.2 2.80 34.3 10,593 140.4 Other 24,993 40.1 2.87 35.1 8,309 110.1 Total 62,310 100.0 8,17 100.0 7,547 100.0 GOVERNMENT FINANCE, consolidated, 1980 US$ Mln. Consolidated Public Sector Receipts 23,588 Expenditures 24,018 Surplus -430 Ratio of Current Receipts to GDP at Market Prices (2) 34.1 MONEY, CREDIT AND PRICES (in billions of diners) 1976 1977 1978 1979 1980 1981 Money Supply 5/ (end-year position) 206.4 251.1 315.3 375.2 461.6 584.3 Money Supply as S of GDP at market prices 30.4 30.1 30.9 29.2 26.7 24.7 Bank Credit Total 6/ 569.1 699.3 959.2 1,217,7 1,568.3 1,927.8 Enterprises 440.4 535.8 728.7 943.0 1,225.1 - Government and other Social Sector 78.2 98.4 140.2 157.6 186.5 households 50.5 65.1 90.3 117.1 156.7 Price Indices (Annual Percentage Change) Industrial Producer Prieas 6 10 9 14 27 45 Agricultural Producer Prices 14 11 13 25 36 54 Retail Prices 10 13 14 22 30 47 1/ The per capita GNP estimate is st market prices, calculated by the same conversion technique *e the World Bank Atlas, All other conversions to dollers in these tables are at the prevailing period *verage exchenge rate. 2/ Includes irrigation and forestry. 3/ Manufacturing, minting, construction, electricity, gas and water. 4/ Total active resident labor force, excluding unemployed. 5/ Currency in circulation, demand deposits and float. 6/ Short- and long-term credits.  - NM - BALAICK OF PAYMENTS. EICUHAIDISC TEA AII DUOT Annual Data at Current Prices 1/ (USS Killions) 1976 1977 1978 1979 1980 1981 SUMKARY OF BALAIICE OF PAYMENTS IMZ UL DElI Deca tr 3L 1980 Exports (f o . ) 4,878 5,254 5,671 6,794 8,978 10,929 US$ Millionm Imports (c.i f ) -7 367 -9,634 -9 98t -14 019 -15 064 -15f 757 Trade Balance 8 4,8 -417 -7,225 ,T6 28 Debt Outstanding and Disbursed 15,435 Official (4,541) Nonfactor Service Receipts 2,051 2,216 2,380 3,771 4,547 5,625 Private (10,894) Nonfactor Service Payments -888 -888 -1.051 -1j624 -1,565 -2 059 Nonfactor Services Net Balance 1,163 1,328 1,329 2,147 2,982 I Debt Service Ratio 3/ (2) 19.0 Factor Services & Transfers Receipts 1,974 2,640 3,120 3,581 4,247 5,927 Factor Services 6 Transfers Payents -838 -1.135 -1 775 -2 504 -3 792 -5 215 Factor Services S Transfers Net Balance 1,136 1,505 1,34 107 455 712 Current Account Balance 165 -1,582 -1,256 -3,661 -2,291 -750 Medium 6 Long-Term Loans 18P1 LltNDING, (as of Sgsember 30, 1981) (million USt) Dimbursments 2,320 2,665 3,150 3,200 3,802 2,790 Amortization -930 -992 -1.326 -1 704 -1 699 1.807 Outstanding & Disbursed 2,154.7 Net Di&burseusnts 1,390 1.673 1,824 1,496 2,103 983 Undisbursed 894.5 Export Credits Extended (net) -100 -213 -105 -150 -215 -234 Capit.l Transactions n.e.i. 2/ -277 172 -204 1,173 553 238 Use of Reserves -1,178 -50 -259 1,142 -150 -237 MERCHANDISE TRADE Imports Capital Goods 1,759 2,436 2,559 3,572 2,939 2,877 Intermediate Goods 4,697 5,989 6,325 8,935 10,630 11,987 Lonsumer Goods 911 1,208 1 099 1,512 1,496 S93 Total Merchandise Imports (c.i.f.) 7,367 9,633 9,983 14,019 15,065 177 Exports C-p.tal Goods 826 1,060 1,129 1,156 1,362 1,736 Intermediate Goods 2,589 2,667 2,809 3.627 4,581 5,361 Consumer Goods 1,463 1,509 1,730 2,011 3,035 3 832 Total Herchandise Exports (f.o.b.) 4,876 5,256 5,668 6,794 B,7 10,129 HERCHANDISE TRADE INDICES 1976-100 Index of Export Dollar Unit Values 100 112 123 141 168 181 Index of Import Dollar Unit Values 100 114 120 143 171 IU Terms of Trade Index 100 98 103 99 98 97 ,;ATE OF EXCHANGE Annual Averages US$1.00 * Dinar 18.19 18.30 18.64 19.0 24.91 35.51 Dinar - USto.055 0.055 0.054 0.053 0.040 0.028 / Valued at statistical exchange rates. 2/ Includes errors and omissions, short-term loans, IMF account, and bilateral balances. Il/ Medium- and lung-term debt service as a percentage of gross current account receipts.  - xvi - YUGOSLAVIA: ADJUSTMENT POLICIES AND DEVELOPMENT PERSPECTIVES Introduction 0.01 By any standards Yugoslavia's performance since World War II has been extremely impressive. In this period a largely rural, peasant economy, shattered by war, has been transformed into a modern urban industrial society with the abolition of absolute poverty and of hunger. This has been achieved despite the difficulties of nation-building in a multinational state with wide inherited regional disparities, while evolving a system of economic management unique to the country. Past World Bank reports have provided an extensive review of Yugoslavia's development achievements and of the economic institutions fashioned by it over this period; for this reason, this account is not repeated here. 1/ Not withstanding this highly successful performance turbulence in the international economy in the 1970s confronted the Yugoslav economy with a new set of challenges, as the economy was obliged to adjust to such developments as worsened terms of trade, higher energy prices, recession in the industrial countries, shifts in international sources of capital and diminished prospects for employment of Yugoslavs abroad. This report uses the occasion of the start of the 1981-85 Yugoslav plan to review the performance of Yugoslavia in responding to these challenges, while maintaining the pace of structural change. The report is written at a time of unusual strain for the Yugoslav economy, when many of the institutions and instruments of economic management are being critically reviewed within Yugoslavia. The report attempts to imake a contribution to this debate. At the same time it should be recognized that policies, targets and institutions are currently in considerable flux, so that elements of this report could be soon overtaken by events. 0.02 This main report is in two parts. The first deals with issues of adjustment strategy and policy at the economywide level, concentrating especially on the period of the 1976-80 medium-term plan. The second part deals with issues in the major sectors of agricultura and industry, and with issues of employment and regional policy. While the energy sector has been of central importance to Yugoslavia's adjustment efforts, it is discussed only in passing in this report as it is being concurrently studies in an energy sector survey being undertaken by the World Bank in cooperation with the Yugoslav authorities. Major policy suggestions are summarized in Chapter X. 1/ See Yugoslavia: Development with Decentralization (Baltimore: Johns Hopkins University Press, 1975) based upon the findings of a World Bank economic mission which visited Yugoslavia in 1972, and Martin Schrenk, Cyrus Ardalan and Nawal A. El-Tatawy, Yugoslavia: Self-Management Socialism - Challenges of Development (Baltimore; Johns Hopkins Press, 1979) based upon a World Bank mission which visited Yugoslavia in November 1976.   PART I: ADJUSTMENT PERFORMANCE AND POLICIES I. THE ADJUSTMENT STRATEGY OF THE 1976-80 PLAN A. Background to the Plan 1.01 Yugoslavia entered the 1976-80 plan period in difficult economic circumstances. These were partly caused by a series of adverse external shocks over the 1974-76 period, and in part reflected longer-term structural tendencies which had been in evidence at least since the economic reforms of the 1965-67 period. In addition the 1976-80 plan was ushered in at a time of substantial change in Yugoslav economic and political institutions. In order to understand both the adjustment strategy of the 1976-80 plan and its subsequent implementation it is useful briefly to review economic and institutional trends in the period that preceded it. Institutional Change 1.02 The Yugoslav constitution of 1963 and a series of economic measures taken between 1964 and 1967 (usually referred to as the economic reforms of 1965) marked a decisive change from earlier Yugoslav systems of economic management. The content of these reforms has been extensively described and commented upon elsewhere 1/ and does not bear detailed recounting here. In brief, though, these measures instituted two concurrent processes: a reduction in the role of the state in economic management ('de-etatization') and a devolution of remaining state involvement from higher-level to lower-level state agencies (decentralization). These processes involved, among other things, the abandonment of centralized investment allocation in favor of autonomous decision taking by commercial banks, the replacement of mandatory plans by a system of indicative planning, considerable devolution of fiscal authority away from the federal level and an increased role for the price system and the market in resource allocation. 1.03 In its actual workings, the "economic model" of 1965 soon displayed certain weaknesses. The devolution of authority over public expenditure and investment decisions deprived federal institutions of the most powerful instruments previously utilized by them to direct resources into areas of social priority. While this role was in theory to have been assumed by market forces, in practice the operation of the market was severely abridged. The mobility of investable funds through the banking system was limited and highly regionalized. The more powerful enterprises were able to preempt investment resources, and the financial accountability of enterprises remained limited. Additional distortions were created by the fact that certain prices remained administratively determined while others were liberalized. As a consequence of these shortcomings certain negative trends in the economy became of growing concern to Yugoslav policy makers in the late 1960s and early 1970s. These included the emergence of considerable 1/ Ibid.  overt inflation, adverse balance of payments trends, and a decline in the efficiency of investment. In addition, there was concern that the system afforded too mucb power to the managerial structures of the society, and thereby weakened the social and ethical goals of self-management. 1.04 Growing awareness of these difficulties, together with other political imperatives, led to further evolution in the system of economic management. The first steps in this direction were amendments to the constitution in 1971. These were followed by a new constitution in 1974 which encompassed these cbanges, introduced others, and prepared the way for major revisions in economic organizations and institutions. Although the motivations and consequences of the 1974 constitution clearly transcended economic issues, a number of the changes were intended as a response to unresolved issues of economic management, and to perceived weaknesses of a 'pure' market system. 1.05 The 1974 constitution attempted to reconcile decentralized decision making with the need for coordinated national action through the development of new procedures, particularly in the area of investment planning. While these procedures admitted the role of the market, this was in a more restricted form than previously and attempts were made to make the operations of the market more consistent with socialism and self-manage- ment. 1/ The new system of planning -- self-management planning -- established procedures for obligatory exchange of information between major agents in the economy (essentially social sector enterprises, financial institutions and government bodies) in an attempt to develop an internally consistent set of investment expenditure and financing plans. The new law on planning of 1976 (which further elaborated the provisions of the constitution) further distinguished between priority and nonpriority activities. Decisions on activities to be accorded priority status were to be arrived at consensually at an early stage in the planning process. In the case of priority activities, consistency of plans by a prescribed date was to be mandatory and implementation of these plans obligatory. The same was not required of plans in the nonpriority sector. 2/ 1.06 The new planning law was only one of a series of legislative measures (mostly passed in the 1974-78 period), designed to bring economic relations in all spheres into harmony with the precepts of the new constitution. As the principles and content of the new legislation have been extensively reviewed in the earlier World Bank study 3/ that discussion is not repeated in this report. However two aspects of the new institutional arrangements merit mention: the internal organization of enterprises, and the provision of nonmarketed services. 1/ See Schrenk et. al., p. 67. 2/ The system of investment planning is described more fully in Chapter IV. 3/ Schrenk, et. al.  - 3 - 1.07 In an effort to reassert genuine democracy in the workplace, enterprises were reorganized into Basic Organizations of Associated Labor (BOALs). These were to be the smallest operating units which produced a marketed or marketable output. The enterprise (or Complex Organization of Associated Labor (COALW) was conceived to be no more than a voluntary association of BOALs, linked together through contractual relations by mutual self-interest. In principle, BOALs were given the rigbt to calve off from the COAL at any future date, subject to conditions agreed upon at the time of association. Detailed regulations were drawn up to govern the calculation of income, its distribution between BOALs, and within BOALs between saving (accumulation) and personal income. In the case of social welfare and other nonmarket services (including such functions as education and health) new organizations called communities of interest (COI) were formed at the republican and local level, to provide these services in their respective jurisdictions. Policy for these bodies was set at the local level and they were to be financed through local levies and contributions. Economic Trends 1.08 The period following the 1965 reform witnessed a substantial shift in the structure of Yugoslavia's balance of payments. As Table 1.1 indicates, between 1965 and 1971 there was a substantial increase in the deficit on merchandise trade, 1/ from about 3% of GDP to almost 10% of GDP. Merchandise exports and merchandise imports contributed roughly equally to this increase in the merchandise trade deficit, although at different times. Between 1965 and 1968 the share of merchandise exports in GDP dropped by four percentage points, but this was partially compensated for by a reduction in the merchandise import ratio. Thereafter the export share stabilized and the import share began to rise, until in 1971 merchandise imports were 22% of GDP. 1.09 As a consequence of these developments there was a sharp decline in the ratio of merchandise exports to merchandise imports, from about 85% in 1965 to 56% in 1971. This trade deficit was offset to some degree by a rising surplus on both nonfactor services and factor services and transfers, the latter item reflecting the growth of workers' remittances which became an important element in the balance of payments after 1965. Despite this offset however, open current account deficits began to emerge; while these were at manageable levels until 1968, the resumption of more rapid growth thereafter precipitated a balance of payments crisis in 1970 and 1971. A package of stabilization measures, which included two devaluations of the dinar, was introduced in 1971. Through adjustment in both exports and imports, and with continued buoyancy in remittances, the situation was turned around in 1972 and a substantial current account surplus of $419 million, equivalent to 2.2% of GDP, was recorded. Slow growth policies were continued in 1973, leading to another surplus, again assisted by a strong performance in remittances. 1/ Unless otberwise indicated, and following Yugoslav practice, the terms "imports" and "exports" in this report refer to trade in goods and nonfactor services, while "merchandise imports" and "merchandise exports" refer to visible trade.  Table 1.1: EXTERNAL SECTOR DEVELOPMENTS, 1965-80 (ratios to current price GDP, in percent) 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 Exports A. Goods 16.3 14.3 13.3 12.3 12.5 12.2 12.3 14.1 14.0 14.2 13.2 13.1 11.5 10.4 10.0 12.9 15.7 B. Nonfactor Services 5.4 5.4 5.5 5.7 6.1 6.3 6.9 8.1 7.8 7.3 6.4 5.5 4.9 4.3 5.6 6.6 7.8 C. Goods and Nonfactor Services 21.7 19.7 18.8 18.0 18.6 18.5 19.2 22.3 21.8 21.5 19.6 18.6 16.4 14.7 15.6 19.5 23.5 Imports D. Goods 19.2 18.5 18.1 17.4 18.1 20.9 22.1 20C5 22.2 28.1 24.9 19.7 21.1 18.2 20.7 21.7 22.6 E. Nonfactor Services 2.2 2.2 2.2 2.1 2.2 2.6 3.5 3.7 3.2 3.1 2.2 2.4 1.9 1.9 2.4 2.3 3.0 F. Goods and Nonfactor Services 21.4 20.7 20.3 19.5 20.3 23.5 25.6 24.2 25.4 31.2 27.1 22.1 23.0 20.1 23.1 24.0 25.6 G. Net Factor Income and Current Tranfers from Abroad ( outflow) -0.3 0.6 0.2 0.4 1.0 2.4 3.5 4.6 6.0 5.3 4.3 4.0 3.2 3.2 2.1 1.2 1.0 H. Balance of Merchandise Trade (A - D) -2.9 -4.2 -4.8 -5.1 -5.6 -8.7 -9.8 -6.3 -8.1 -13.9 -11.7 -6.7 -9.6 -7.9 -10.7 -8.8 -6.9 I. Resource Gap (C - F) 0.3 -1.0 -1.5 -1.5 -1.7 -5.0 -6.3 -2.0 -3.6 -9.7 -7.6 -3.6 -6.7 -5.5 -7.5 -4.5 -2.1 J. Current Account Balance (G + I) 0.0 -0.4 -1.3 -1.1 -0.7 -2.6 -2.4 2.7 2.4 -4.4 -3.2 0.4 -3.5 -2.3 -5.4 -3.3 -1.1 Memo: Percentage Growth in GDP (Constant 1972 prices) 1.4 5.0 3.3 5.8 11.2 5.0 8.8 3.6 2.6 14.7 0.9 5.3 8.5 8.5 4.2 2.4 2.2 Percentage Change in Terms of Trade 1.9 2.8 0.0 -2.7 -1.0 1.9 1.0 -0.2 -0.7 -6.0 1.9 0.5 -0.7 3.4 -2.5 -0.5 -1.3 Current Account Balance (S mn.) 73 -33 -82 -106 -110 -340 -357 419 485 -1,183 -1,003 165 -1,582 -1,256 -3,661 -2,291 -750 1/ Trade values taken from dollar balance of payments data, converted at period average exchange rates. Source: Statistical Appendix Tables A 2.1, A 3.1, A 3.2; Statistical Yearbook, 1981, Table 102.27; World Bank Report No. 2972-YU, Export Performance and Prices, October 16, 1980, Table A.1.  - 5 - 1.10 Following the success of this stabilization episode, macroeconomic policy became more expansionary in 1974, and growth proceeded extremely rapidly. however the timing of the boom coincided witb the increase in oil prices and the recession in the industrial market economies. These developments had a substantial effect on Yugoslavia's terms of trade and on the growth of principal markets for Yugoslavia's exports, and checked the growth in workers' remittances. There was consequently a further large increase in the merchandise import ratio (in nominal terms) with equivalent increases in the trade gap and the current account deficit. Another stabilization effort had to be mounted, yet again sharply curtailing growth, and restricting the growth of import volumes. 1.11 By the beginning of the 1976-80 plan therefore a stop-go pattern of growth had become an established feature of Yugoslav economic performance. The repetitive nature of these cycles had led Yugoslav economists to refer to them as the 'trade scissors' cycles of domestic economic growth. 1/ The beginning of each such cycle was marked by a significant expansion of exports and a sharp increase in the ratio of export growth to import growth. As domestic economic growth picked up, import growtb quickened and export growth declined, reducing the ratio to a critical point that marked the turning point of the cycle. At this point economic growth rates began to fall as a consequence of foreign exchange shortages that interrupted the flow of imported inputs, and as a result of policies designed to slow the economy. This slowdown generated a growth in exports and a reduction in imports which permitted policy to become more expansionary, and growth to resume. Underlying this cyclical pattern was the shift in the structure of the balance of payments that had occurred. This structure consisted of a large deficit in merchandise trade, about 12% of GDP, covered by a relatively predictable surplus on nonfactor services and a substantial but uncertain surplus on factor service account. This was rightly regarded by the Yugoslav authorities as a weak structure, and one likely to constrain growth in the future. Accordingly improvement in the external sector was highlighted as a major structural objective of the 1976-80 Federal plan. 1/ The concept of 'trade scissors cycles' was first introduced by Horvat in his book on business cycles. See Branko Horvat, Business Cycles in Yugoslavia, International Arts and Science Press, 1971. More recently, a consortium of Yugoslav economists has demonstrated the continued relevance of this concept to the cyclical fluctuations of the 1970s. See Mladjen Kovacevic, Ljubomir Madzar, Zvonimir Marovic, Sofija Popov, Zoran Popov, Davor Savin and Nikola Zelic, "Basic Determinants of and Conditions for Dynamic Development and Balanced Economic Trends in the 1981-85 Period," Institute of Economic Sciences, Belgrade, 1980.  - 6 - B. The Adjustment-Strategy 1.12 The 1976-80 plan envisaged external adjustment occuring through a combination of export promotion and import substitution, while maintaining bigh growth rates of output, investment and employment. The major foreign trade and output targets of the plan are presented in Table 1.2. It is clear from this table that, by the standards of the past, the plan's adjustment strategy was heavily weighted toward import substitution. The plan called for 8% growth in exports. This was in fact slightly lower than the actual rate of 8.4% realised in the 1971-75 period and did not seem to imply any radical shift in course. On the other hand the projected growth in imports of 4.5% did represent a substantial break witb the past. When taken in conjunction with the target growth rate of 7% in gross material product (GMP) this import target implied an aggregate elasticity of 0.6, well below either the actual level of 1.5 in the 1971-75 period or the longer-term level of around 1.1 recorded in the 1960s. 1/ Table 1.2: EXTERNAL SECTOR TARGETS OF THE 1976-80 PLAN Actuals 1/ Plan -- - 1971-75 - -1976-80 Real Growth Rates (x) Gross Material Product 5.9 7.0 Exports (Goods and Nonfactor Services) 8.4 8.0 Imports (Goods and Nonfactor Services) 8.8 4.5 Fixed Capital Formation 7.1 8.0 Elasticities (wrt GMP) Exports (GNFS) 1.4 1.1 Imports (GNFS) 1.5 0.6 1/ Figures are calculated point-to-point. As such they differ from data presented in Schrenk, et. al. which are calculated from three year averages. Source: Schrenk, et. al., Table 9.17; Statistical Appendix Tables A.3.1 and A.3.2; material provided by the Yugoslav authorities. 1/ See Schrenk et. al., Tables 9.8 and 9.17.  - 7 - 1.13 The choice of an adjustment path heavily slanted toward import substitution was influenced by several perceptions. There was, firstly, concern at the rapid increase in the economy's import dependence following the post-1965 liberalization. A rise in import dependence over this period was consistent with the principal objective of the 1965 reforms: opening the Yugoslav economy to external trade, and aligning the productive structure more closely with international comparative advantage. However, Yugoslav policymakers were also influenced by a long-standing and widely-held belief, that internal price controls, coupled with the structure of protection had resulted in a distorted pattern of relative prices domestically. As a result of these purported distortions, production in many raw material and basic industries was thought to have lagged behind levels that were economically rational, even for an economy more responsive to international cost patterns. Concern over increased dependence on intermediate imports was aggravated by world trends in energy prices and the growing burden of payments for imported oil in the total import bill. Changes in world energy prices were thought to warrant a substantial effort to develop domestic energy resources, and to restructure domestic plant and equipment in order to reduce dependence on foreign oil. At the same time the boom in international commodity prices in the early 1970e provided added incentive to development of Yugoslavia's mineral resources. 1.14 These arguments in favor of economically efficient import substitution were coupled with some pessimism on the prospects for growth in exports mucb above the historical trend rate. It was felt that export possibilities were limited by high technology competition and protectionist barriers in the developed market economies, and by low wage competition in labor intensive products from lesser developed countries. Equally, the oscillations in prices and output in the international economy of the early seventies had buffeted Yugoslavia along with other developing countries, and an import substituting strategy was seen as offering a more certain path to external balance, while making the economy less vulnerable to external shocks. Adding support to this export pessimism were developments at the time of preparation of the plan, including the growth of barriers against Yugoslav agricultural exports by the EEC in 1974 and the recession in the developed market economies at that time. It should be noted that Yugoslavia was not alone in choosing an import substitution strategy as the dominant mode of adjustment to the external shocks of the 1974-78 period: this was the dominant tendency among the semi-industrial developing countries. 1/ 1.15 In general support of these aggregate import substitution goals, but more specifically to address the structural imbalances that were perceived to exist in the domestic productive structure, the 1976-1980 plan called for an ambitious restructuring of the economy. This was to be achieved tbrough a substantial investment effort consciously directed at 1/ See World Development Report, 1981, Chapter VI.  - 8 - designated priority sectors. These priority sectors were to receive preferential access to investment funds in order to permit them to grow more rapidly than the rest of the economy. The priority sectors were: electrical energy, coal extraction, oil and gas extraction and refining, ferrous metals, nonferrous metals, manufacture of basic chemicals, extraction of nonmetallic minerals, machinery and shipbuilding, agroindustry, highway transportation and foreign tourism. 1.16 Investment in fixed assets was planned to grow at 8% a year over the plan period, up from an already substantial growth rate of 7.1% between 1971 and 1975. Of this investment 61% was to occur in the so-called 'economic' sectors; the remainder in the noneconomic sectors 1/. The aim of the priority allocation mechanism was to influence the distribution of this economic fixed investment between branches. As the data in Table 1.3 indi- cate, the share of priority sectors was slated to increase from 53.3% of total economic investment to 64.6% in the 1976-80 period. These figures attest to the magnitude of the restructuring of investment foreseen in the plan. As the table indicates, the major shifts envisaged were in the electric power, oil and gas and chemicals sectors, with continued priority accorded to highway transportation. The investment strategy was thus clearly designed to strengthen the raw material and infrastructure base of the economy. By contrast, with the exception of the machinery and shipbuilding sectors, there appears to have been little attention paid to the creation of capacities explicitly for export. 1/ Yugoslav statistical conventions divides economic and social activity into the 'economic' and 'noneconomic' sphere, where the 'economic' sphere covers the production of goods and services for commercial purposes, and the 'noneconomic' sphere covers such social services as education, health, housing and administration. For national income accounting, Yugoslavia follows a somewhat different classification scheme. In line with Marxian concepts, income is considered to be generated only by 'productive' activities. These include all economic sectors which produce gooda, but only a portion of the economic sectors which produce services, together with a portion of the noneconomic sectors which generate services of use in national production. Output from economic activities is included in the Yugoslav measure of national output, gross material product (GMP; sometimes referred to as gross social product (GSP)). The principal nonproductive branches are health, education, administration, defense, banking and housing. GNP and GDP figures in this report have been estimated by the World Bank on the basis of Yugoslav data on GMP, and additional statistical information on the service sectors excluded from GMP. For additional details please see Appendix IV.  - 9 - Table 1.3: STRUCTURE OF INVESTMENT IN FIXED ASSETS IN ECONOMIC SECTORS, 1971-75 and 1976-80 (in percent; underlying data based on constant price estimates, 1975=100) 1976-80 1971-75 (actual) 1976-80 Plan (Estimate) All Priority-Sectors 53.3 64.6 59.5 of which: Electrical Energy 9.4 12.1 14.9 Coal Extraction 1.3 1.8 2.0 Oil & Gas Extraction 1.8 6.1 4.8 & Refining Ferrous Metals 3.2 3.6 3.3 Nonferrous Metals 3.2 3.6 4.2 Manufacture of Basic 1.8 4.6 3.6 Chemicals Extraction of Nonmetallic 0.3 1.1 0.7 Minerals Machinery & Shipbuilding 4.0 2.8 4.3 Agroindustry 9.9 11.0 9.2 Highway Transportation 14.4 14.2 10.1 Foreign Tourism 4.0 3.7 2.4 Nonpriority Sectors 46.7 35.4 40.5 Total Fixed Investment (Productive Sectors) 100.0 100.0 100.0 Source: Data provided by the Yugoslav authorities. 1.17 The desired acceleration in fixed investment translated itself into a faster growth in aggregate investment over the period, 7.2% instead of 5.3% in the 1971-75 period. Given a target growth rate in gross material product (and by assumption, an equivalent growth in gross domestic product), this implied a virtually unaltered investment share in GDP, of about 29%. The plan documents themselves did not provide an explicit financing pattern for this investment effort in national accounts terms. The 1976 Bank mission did however make estimates of the resource balance implied by the plan's foreign trade targets, and of the net support likely to be available from  - 10 - factor service receipts, particularly net workers' remittances. The results of its analysis are shown in Table 1.4. As can be seen, the domestic and national savings rates implied by the 1976-80 plan on this reckoning were within the limits of past experience. Table 1.4: MACROECONOMIC INDICATORS 1971-75 (Actual) AND 1976-80 (Planned) (as ratios to GDP) 1971-75 1976-80 Gross Investment Rate 29.8 29.2 National Saving Rate 27.6 27.1 Domestic Saving Rate 24.0 24.9 Source: Schrenk, et. al., Table 5.6 1.18 There remained the further issue of the availability of external capital to finance the gap between gross domestic investment and gross national savings. The external financing requirements for the 1976-80 period were not explicitly articulated in the plan documents reviewed by the 1976 mission. However the plan did establish a cumulative current account deficit of $5.0 billion over the five year period, in nominal terms. Taking into account projected amortization of debt existing at the beginning of 1976 and the need to maintain adequate reserve cover, this target current account deficit implied gross external borrowing of $9.8 billion over the period. The implied increase in medium- and long-term (MLT) external debt outstanding and disbursed (DOD) over the period was $6.0 billion, as compared to a stock of MLT DOD of $5.8 billion at the end of 1975. On the basis of a similar analysis of capital requirements, the 1976 mission concluded that adequate external financing was likely to be available to Yugoslavia without any significant increase in the debt service ratio.  - 11 - Table 1.5: ESTIMATED EXTERNAL FINANCING REQUIREMENTS OF THE 1976-80 PLAN (in billions of US Dollars) Projected Current Account Deficit 5.0 Projected Increase in Reserves 1.0 Scheduled Amortization 3.8 Gross Medium- and Long-Term Borrowing 9.8 Net Increase in Medium- and Long-Term Debt 6.0 Source: World Bank estimates. 1.19 Taken overall, and given the circumstances of the time, the plan represented a credible and consistent scenario for the adjustment needs that the country faced at the end of 1975. The strategy in many respects accorded with the adjustment profile recommended by the World Bank 1/ of high investment and savings, and only a temporary interruption in growth. As the Bank has noted, "programs to expand the production of energy and tradeable goods require substantial new investment. /In addition/ a high investment economy is able to "turn over" its old capital stock quickly so as to reflect new scarcities, especially more expensive energy". 2/ Given Yugoslavia's already high savings and investment rate the need for additional savings was perhaps lower than that in otber countries. The main criticism that could have been made of the strategy at the time was that it was excessively biased toward import substitution goals. As discussed in detail below, there was already evidence of faltering export performance at the time the plan was being prepared. While the import substitution bias of the plan was both acknowledged and endorsed by the 1976 World Bank mission, it was pointed out that the targeted growth of output and imports projected in the plan were insufficient for the growth in aggregate final demand that was projected. 1.20 The 1976-80 plan was thus explicitly geared toward external adjustment and a conscientious effort was made to translate its strategy into practice. Despite these efforts, by the end of the plan Yugoslavia was suffering its most intense foreign exchange crisis in two decades, accompanied by unprecedented inflation and the prospect of sustained low growth rates for five years in a row. The deterioration in Yugoslav economic performance that became evident in the post-1979 period raises questions about the factors contributing to the crisis. How did Yugoslavia move from a current account surplus of $165 million in 1976 to a deficit of $3.7 billion in four years? Was the crisis the outcome of an inappropriate strategy or of 1/ See World Development Report, 1981, Chapter VI. 2/ Ibid., page 66.  - 12 - failures of implementation and policy, as its resemblance to past foreign excbange crises would indicate, or can it be explained by exogenous shocks coming from the world economy? Answers to these questions are not just of interest in themselves but are of importance in assessing the strategy of the 1981-85 plan, and of the policy framework being used to implement it. 1.21 The analysis in Part I of the report addresses these and related questions about the outcome of the 1976-80 plan and the origins of the present crisis, as a prelude to an assessment of the present adjustment strategy. In both the historical and the forward-looking discussions the analysis draws extensively upon the simulations of a computable general equilibrium (CGE) model of Yugoslavia constructed for this report. The model is described in Appendix I. The next two chapters review plan performance in the key areas of the balance of payments, output, investment and growth. Chapter IV reviews the policy environment in the 1976-80 period, particularly as it influenced the allocation and pricing of capital and of foreign exchange, and draws upon this analysis to provide a summary appraisal of the experience of the 1976-80 period. Chapter V provides an assessment of the adjustment strategy of the 1981-85 plan, and of developments whicb have occurred since the 1981-85 plan was launched.  - 13 - II. PLAN PERFORMANCE: OVERVIEW AND EXTERNAL SECTOR DEVELOPMENTS A. An Overview 2.01 The cyclical relationships between imports, exports and domestic growth characteristic of Yugoslavia in the late sixties and early seventies persisted in the 1976-80 period. Major indicators for the period are provided in Table 2.1. The economy began the period in a trough caused by policy measures designed to ease the foreign exchange crisis of 1974-75. As a result, output growth in 1976 was the lowest realized in Yugoslavia in the postwar period. Depressed domestic conditions, combined with severe restrictions on imports, allowed for a small current account surplus and a reserves buildup. This gave the authorities the confidence to proceed witb the more expansive strategy of the five-year plan. 2.02 Growth proceeded rapidly in 1977, powered by a more liberal monetary and import policy. Fixed capital formation grew by 12% after a cutback the previous year, and the real import of goods and nonfactor services rose by 12.2%. In response to the booming domestic economy, export growth slackened significantly; this, combined with fast import growth led again to a large current account deficit, of $1.3 billion, and to an acceleration in inflation. The strength of this expansion was regarded as unsustainable and policy became mildly restrictive in 1978. There was a reduction in botb imports and investment expenditures. The overall growth rate declined slightly but remained at the relatively high level of 6.8%. The slowdown in growth did not appreciably affect the inflation rate; nevertbeless there was some improvement in the growth rate of exports. 2.03 The boom in the domestic economy survived into 1979. However the combination of rising domestic inflation, deteriorating balance of payments trends and the continuing sluggishness of growth in the OECD area made it increasingly clear that the expansion could not be sustained. Three additional shocks considerably aggravated the situation. These were: a poor harvest in 1978 which created a need for substantial agricultural imports in 1979; an earthquake in Montenegro which adversely affected tourism receipts in the affected area (and perhaps more widely) and the increase in the price of oil imports, which increased Yugoslavia's expenditure on imported crude oil by about $500 million. 2.04 The combination of these events led to a current account deficit of $3.7 billion in 1979, equivalent to 5.4% of GDP, the largest deficit in relative terms that Yugoslavia had sustained since the mid-sixties (see Table 1.1). A stabilization program was instituted toward the end of 1979 consisting of tight monetary policy and direct controls on investment expenditures, particularly in the noneconomic sector. This program was supported by a standby arrangement with the IMF, concluded in May 1980. In June 1980 the dinar was devalued by 30% against the U.S. dollar in gross terms. 1/ In addition imports were restricted to essential items. 1/ The devaluation coincided with the abolition of a 10% import surcharge, coupled with some reduction in export rebates. As a result the net devaluation was somewhat less than this.  - 14 - Table 2.1; MAIN INDICATORS OF THE 1976-80 PERIOD 1976-80 -1976 - 1977- 1978 1979 1980 Actual Plan 1/ Real Growth Rates in Percent (value data deflated to 1982 prices) Gross Material Product 3.9 8.0 6.8 7.0 2.2 5.6 6.9 Social Sector Employment 3.6 4.5 4.5 4.3 3.2 4.0 3.5 Industrial Production 4.0 9.4 8.7 8.3 4.4 6.9 8.0 Agricultural Output 7.0 5.0 -6.0 6.0 -2.0 1.9 3/ 4.0 3/ Exports of Goods and Nonfactor Services 2/ 10.7 -4.4 -0.2 12.6 7.7 5.3 7.8 Merchandise Exports 15.2 -4.2 -1.4 4.4 11.2 5.0 7.0 Imports of Goods and Nonfactor Services 2/ -4.4 12.2 0.6 18.5 -11.1 3.2 4.1 Merchandise Imports -6.9 15.1 0.6 17.3 -10.1 3.0 3.8 Fixed Capital Formation 7.5 11.7 13.4 4.6 -1.5 8.1 7.0 Nominal GrowthbRates, in percent Retail Prices 9.4 13.3 13.4 21.9 29.9 17.4 - Millions of Current Dollars Balance of Payments: Current Account = Deficit) 165 -1,582 -1,256 -3,661 -2,291 -8,625 -5,000 1/ Target growth rates cited bere are based on the actual outturn in 1975, while elsewhere in the text target growth rates are based on estimates of 1975 performance at the time that the 1976-80 plan was prepared. 2/ There are several problems with Yugoslav current price balance of payments data, particularly in the use of "statistical" rather than market exchange rates in aggregating trade denominated in different currencies. In addition, relatively small variations in deflators can make for larger differences in estimated real rates; the real estimates presented above should therefore be treated with some caution. 3/ Base = 1974/5 Average. Source: Statistical Appendix Tables (various) and information provided by the Yugoslav authorities.  - 15 - 2.05 These measures had a substantial effect on both growth and the balance of payments. Growth of material product in 1980 declined to 2.2%, a level lower than that of 1976. Imports were sharply cut while exports grew substantially. As a result the deficit on merchandise trade was reduced by $1.2 billion. This, coupled with a modest increase in the surplus on services led to a substantial improvement in the current account deficit, to $2.3 billion or 3.3% of GDP. The shift of resources to the external sector was made possible by a cut in fixed capital formation. In addition the growth in consumption was sharply curbed primarily through a fall in real wages in the social sector. Despite this drastic slowdown inflation accelerated. In part this reflected the impact on the domestic economy of the increased price of imported energy and the devaluation of the dinar. In addition, however, as analysed below (Chapter VIII), it represented the attempt of enterprises to rebuild their liquidity and savings in the face of restricted credit availability. 2.06 The consequences of these current account developments were felt by the capital account. There was a rapid growth in external debt, both medium- and long-term and short-term. In contrast to the planned increase in medium- and long term debt outstanding and disbursed of $6 billion over the period (cf. para 1.18), the actual increase was $9.6 billion, and the buildup of short-term debt and bilateral payment obligations was over and above this. This growth in debt in turn affected the current account by imposing additional interest payment obligations. As the incremental debt was increasingly on a floating rate basis, Yugoslavia became even more vulnerable to the increase in Eurodollar interest rates which occurred after 1979. Gross interest payments increased from $0.4 billion in 1976 to $1.3 billion in 1980 (Statistical Appendix Table A 3.2). 2.07 Taking the plan overall, in substantial measure Yugoslavia succeeded in achieving its aggregate growth and investment objectives. Yet at the end of the plan its external position was considerably more vulnerable than had been the case at the beginning. As is evident from the above narrative, external shocks were to a considerable degree responsible for this outcome. The remaining sections of this chapter and the next two chapters attempt to assess the extent to whicb domestic performance and policies could have mitigated the force of these external shocks, and also the degree to which the underlying structural objectives of the plan were fulfilled despite the shocks encountered from the outside world. B. The External Sector Sources of Change in the Current Account Deficit 2.08 Since management of current account deficits was the prime preoccupation of Yugoslav policymakers over the plan, it is of interest to decompose the influences on the current account deficit, both to identify the factors of greatest quantitative significance, and to distinguish between exogenous shocks and elements more within the domain of Yugoslav policies. The analysis of Table 2.2 takes the current account deficit incurred in 1975 (of $1.0 billion) as a benchmark, and compares the current account deficit of each year in the plan with reference to this benchmark.  - 16 - 2.09 Three sets of effects are identified: real factors, price effects and the nominal impact of increased interest payments. In the first block, 'real factors', exports and imports are defined to include all goods and services, other than interest receipts and payments. These flows are valued at 1975 prices to measure changes in real resource transfer as compared to 1975. The contribution of price movements to the deficit is isolated in the second block, which combines the effects of general inflation in traded goods over the period and the effects of terms of trade changes. It should be emphasized that this decomposition is relatively crude, in that it does not attempt to capture important interactions between the individual elements. The impact of price developments on the deficit for Table 2.2: SOURCES OF CHANGE IN CURRENT ACCOUNT DEFICIT, 1975-80 (US$ million) 1976/75 1977/75 1978/75 1979/75 1980/75 Change in Nominal Deficit (over 1975 deficit) 1168 -579 -253 -2658 -1288 A. Real Factors 1/ 880 -599 -591 -1656 -255 Exports (+ = increase) 530 128 476 1218 1526 Imports (- = Increase) 350 -727 -1067 -2874 -1781 B. Price Effects 2/ 292 3 363 -644 -224 C. Interest Payments (net) -4 17 -25 -358 -809 Receipts 28 61 93 126 135 Expenditures -32 -44 -118 -484 -944 1975 1976 1977 1978 1979 1980 Memo: Current Account Balance -1003 165 -1582 -1256 -3661 -2291 1/ Exports and Imports refer to exports and imports of all goods and services with the exception of interest payments. They are deflated by their respective merchandise trade indices. 2/ Residually derived. Source; Statistical Appendix Tables A 3.1, A 3.2, A 3.5 and A 3.6; Mission Estimates  - 17 - example clearly depends upon the size of the underlying real deficit. More complex decomposition schemes which take such interactions more explicitly into account have also been devised. 1/ 2.10 It is evident from this table that roughly two-thirds of the $2.6 billion increase in the current account deficit from $1.0 billion in 1975 to $3.7 billion in 1979 was attributable to 'real' factors, with price factors accounting for a further quarter of the deterioration, and increased interest payments for the remaining one-sixth. By 1980 however, the pattern was significantly different. As a consequence of deep cuts in imports and a successful export drive, the contribution of 'real' factors to the increase in the deficit had been almost wholly eliminated and the increase in the deficit as compared to 1975 was due to price effects and increased interest expenditures, in roughly equal measure. The table thus reveals the 1979 crisis to be the outcome of three interrelated events: a big increase in real imports not matched by exports, thereby widening the trade deficit, adverse price movements which considerably compounded the financial consequences of this widened real deficit, and a sharp increase in interest payments on account both of rising rates of interest, and of the increased borrowing needed to finance the deficit. 2.11 Although price developments, both directly and indirectly (through their impact on the need for additional borrowing and on the nominal level of international interest rates) were an important source of the 1979 crisis, their quantitative impact was as sizeable as it was primarily because of the large real trade deficit in that year. Table 2.3 examines the evolution of this deficit by components, in real terms, using the merchandise deflators for the various components (including services) for comparability with Table 2.2. The growth in the 'real' deficit between 1975 and 1979 is seen to reflect several developments. After a strong spurt in 1976 (a year of domestic recession), export growth slowed across the board in 1977 and 1978, although there were additional secular influences, such as a sharp decline in the exports of the shipbuilding sector, one of the major exporting sectors at the beginning of the plan. The real surplus on nonfactor services stagnated, reflecting both weak demand for tourism and transport services in the early years of the plan, and strong growth in Yugoslav expenditures on transportation services as Yugoslav trade was increasingly carried on ships of foreign registry. Finally, as had been expected, net remittances declined in real terms with the reduction in the overseas work force. 2.12 The syndrome that had characterized the Yugoslav balance of payments in the first half of the 1970s thus persisted into the second half as well. Against the background of sluggish growth in exports, a declining real surplus on workers' remittances, and a stagnating real surplus on nonfactor services, real import growth had to be held in check throughout the period: between 1975 and 1979 merchandise imports grew at 5.7% per year, and declined from 25% to 21% of GDP. The balance of payments was thus 1/ See Bela Balassa, The Newly Industrializing Developing Countries After the Oil Crisis, World Bank Staff Working Paper, No. 437, October 1980 and World-Development Report, 1981, Chapter 6.  - 18 - Table 2.3; EVOLUTION OF THE REAL DEFICIT, 1975-80 1/ (in 1975 prices) Average Annual Growth 1975 1976 1977 1978 1979 1980 1976-80 Merchandise Exports 4,072 4,690 4,491 4,430 4,622 5,130 5.0 Merchandise Imports -7,694 -7,155 -8,234 -8,189 -9,602 -8,608 2.9 Trade Balance -3,622 -2,464 -3,743 -3,759 -4,980 -3,478 Nonfactor Service Exports 1,982 1,992 1,894 1,951 2,583 2,598 6.3 Nonfactor Service Imports -695 -862 -759 -862 -1,112 -894 7.0 Nonfactor Service Balance 1,287 1,130 1,135 1,090 1,471 1,704 Workers' Remittances (Inflows) 2,004 1,906 1,801 2,153 2,071 1,856 -1.0 Workers' Remittances (Outflows) 2/ -397 -419 -520 -803 -946 -1,065 22.9 Workers' Remittances Balance 1,607 1,487 1,281 1,351 1,125 791 Balance on Goods, Nonfactor Services and Workers' Remittances -728 152 -1,327 -1,319 -2,384 -983 Memo. Export Price Index 100 104.0 117.0 128.0 147.0 175.0 Import Price Index 100 103.0 117.0 122.0 146.0 175.0 Terms of Trade 100 101.0 100.0 104.9 100.7 100.0 Growth over Previous Year: Merchandise Exports 15.2 -4.2 1.4 4.3 10.9 Merchandise Imports -7.0 15.1 0.5 17.3 -10.4 1/ All receipts deflated by export price index, all outflows by import price index. 2/ Recorded under nonfactor service outflows in 1975. Source: Statistical Appendix Tables A3.1, A3.2, A3.5 and A3.6.  - 19 - at best in a somewhat uneasy balance at the beginning of 1979. The various shocks of 1979 destroyed this balance by simultaneously imposing price shocks, interest rate shocks and an additional need for imports on the economy. The pattern of response, rather as in 1976, was again a reduction in domestic growth and imports and a renewed spurt in exports. This had the effect of eliminating the real deficit that had emerged between 1975 and 1979; the difference between the current account deficit of $2.3 billion in 1980 and $1.0 billion in 1975 thus entirely reflected the effects of inflation in traded goods prices, terms of trade changes and increased interest payments. 2.13 From the above analysis it is clear that Yugoslavia had not been able to reorient its balance of payments by the time that the shocks of 1979 occurred, and that the bulk of the difficulty continued to arise from exports, particularly merchandise exports. In order to understand better the factors which determined Yugoslavia's export performance in this period, a more detailed look is now taken at Yugoslavia's export performance over the longer term. This analysis draws in part on an earlier World Bank report (Yugoslavia: Export Performance and Policies, Report No. 2972-YU). 1/ In view of the detailed coverage provided in that report only the major components of the Yugoslav export structure are discussed here. In addition export performance in industry and agriculture are further discussed in Chapters VI and VII. Given the importance of the import regime and structure for the pattern of incentives in the economy, import behavior in the period is discussed later, in Chapter III. Merchandise Exports 2.14 The report cited above documented the fact that Yugoslavia was a strong exporter in the early sixties, with export growth faster than the growth of world trade. However, after 1966 Yugoslavia's export effort entered a period of slower growth and Yugoslavia lost its share of world exports. Concurrent with this slowdown in aggregate exports were significant changes in commodity and market structure. Since the early 1950s diversification of exports by market area has been an explicit objective of Yugoslav policy, with three broad market areas being distinguished: the industrial market economies of the OECD (DCs), the developing countries (LDCs) and the centrally planned economies (CPEs). Tables 2.4 and 2.5 provide information on the market and commodity classification of Yugoslav exports across these trading groups in the 1970s. 2/ 1/ Restricted circulation. For internal World Bank use only. 2/ The market categories used in Table 2.4 in general conform to definitions used in the World Bank's World Development Report. In particular, this implies that the countries of Southern Europe are classified as developing countries. Following the Bank's standard taxonomy, the group of oil exporting developing countries includes both members of OPEC and thirteen non-OPEC members. The EEC is defined to include the nine present members consistently throughout the analysis. For further details please see Report 2972-YU, Annex I. As these data are assembled from partner country sources they do not necessarily match export data from Yugoslav sources presented elsewhere in this report.  1' Tabl 2.4: AGGRIICAT7 REIONAL AND COMMODITY DIlTRIBUTION OF EXPORTS 1970, 1975 AID 1979 veorts (I_nes is illSon 1US$) Regional Distribution (t) Coodity Distribhtion (X) Of Which Centrally Of which Centrally Of AhIOh Centrally Developed Of which DI-lopleg Oil Pla.ned Developed Of whifh Developing Oil Planned Developed Of which Developing Oil Planned ConeItrie _ Conries gortine U-e.In World Co.ntrtes EbC Countries Enrorotie tonofio World Countries EEC Cowntrien Enortiew Zeon8 World 1970 Prlury 508.6 394.5 63.9 16.4 177.5 750.0 67.8 52.6 8.5 2.2 23.7 100.0 56.8 60.2 26.8 24.0 32.6 64.7 AgrSieltt 304.4 243.2 38.7 8.6 88.6 431.6 70.5 56.3 9.0 2.0 20.5 100.0 34.0 37.1 16.2 12.6 16.2 25.7 metals 6 Mlnerals 204.2 151.3 25.2 7.8 88.9 318.3 64.1 47.5 7.9 2.5 27.9 100.0 22.8 23.1 10.6 11.4 16.3 19.0 gacefaerars 386.8 260.9 174.6 51.9 367.6 929.2 41.6 28.1 18.9 5.6 39.6 100.0 43.2 39.8 73.2 75.9 67.4 55.3 Chiseloa 31.9 22.2 13.4 4.6 51.9 97.1 32.8 22.9 13.6 4.7 53.4 ]&2.Q 3.6 3.4 5.6 6.7 9.5 5.8 Machine & Traenpnrtaticn Equip_nt 110.2 78.1 112.3 27.9 158.8 381.3 28.9 20.5 29.4 7.3 41.7 100.0 12.3 11.9 47.1 40.8 29.1 22.7 Other Meeefaetee 244.7 160.6 48.9 19.5 157.2 450.8 54.3 35.6 10.9 4.3 34.9 100.0 27.3 24.5 20.5 28.5 28.8 26.8 Totl 895.3 655.4 238.5 68.4 545.3 1.679.1 53.3 39.0 14.2 4.1 32.5 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1975 m_ry 696.5 454.5 135 1 75.9 539.4 !,371.0 50.8 33.2 9.9 5.5 39.3 100.0 50.3 48.9 17.5 23.4 28.1 33.7 Agrlooltor 441.1 327.2 96.7 51.6 150.3 688.1 64.1 47.6 14.1 7.5 21.8 100.0 31.9 35.2 12.5 15.9 7.8 16.9 Me tls & Minerals 255.4 127.3 38.4 24.3 389.1 682.9 37.4 18.6 5.6 3.6 57.0 100.0 18.4 13.7 5.0 7.5 20.3 16.8 Man.ofaotetue 686.9 475.4 635.5 249.0 1,378.8 2,701.2 25.4 17.6 23.5 9.2 51.0 100.0 49.7 31.1 82.5 76.6 71.9 66.3 Ch-itole 77.0 65.5 60.1 45.4 242.2 379.4 20.3 17.3 15.8 12.0 63.8 100.0 5.6 7.0 7.8 14.0 12.6 9.3 Machinery & Trensporteticn gqeipvaant 234.7 183.2 409.8 101.0 498.0 1,141.5 20.6 16.1 35.8 8.8 43.6 100.0 17.0 19.7 53.1 31.1 26.0 28.0 Other mesufattree 375.2 226.7 166.6 102.6 638.6 1,180.3 31.8 19.2 14.1 8.7 54.1 100.0 27.1 24.4 21.6 31.5 33.3 29.0 Tota1 1.383.4 929.9 770.6 325.0 1.918.2 4.072.2 34.0 22.8 18.9 8.0 47.1 100.0 100.0 100.0 100.0 100.0 100.0 L0 0O 1979 Priesry 1,259.1 855.1 526.0 131.6 702.9 2,288.0 55.0 37.4 14.2 5.9 30.7 100.0 44.0 42.6 23.9 16.2 24.7 32.3 Agronoltwra 811.7 614.1 269.6 109.1 218.7 1,300.0 62.4 47.2 20.7 8.4 16.8 100.0 28.4 30.6 19.8 13.4 7.7 18.3 Metale & min-rele 647.4 241.1 56.4 22.5 484.2 988.0 45.3 24.4 5.7 2.3 49.0 100.0 15.6 12.0 4.1 2.8 17.0 14.0 Mlnofacteree 1,603.1 1,151.9 1,035.6 681.4 2,146.5 4,785.2 33.5 24.1 21.6 14.2 44.9 100.0 56.6 57.4 76.1 83.6 75.3 67.7 Chde1ls 172.7 123.3 165.0 86.6 325.9 663.6 26.0 18.6 24.9 12.1 49.1 100.0 6.0 6.1 12.1 9.9 11.4 9.4 Machinery & Transportation 4eivP-t 537.9 464.7 460.3 303.2 1,101.9 2,100.0 25.6 22.1 21.9 14A 52.5 100.0 18.9 23.2 33.9 37.3 38.7 29.7 Other laIefaotnra 892.7 563.9 410.3 297.7 719.7 2,021.6 64.2 27.9 20.3 14.7 35.5 100&L 31.2 28.1 30.2 36.6 25.2 28.6 Tetal 2.S62.3 2.007.0 1.361.6 913.0 2.849.3 7.073.2 40.5 28.4 19.3 11.5 40.3 100.0 100.0 100.0 100.0 100.0 100.0 7.00. 1/ Casdity gre ar` dfid as fellos (SITC SetSone): Agrlneltwrn: 0+1+2 - (27+26)+4 Metals & MI-nls: 3+27+2867+68 Chedeal: 5 MahieSary & Trnportaton 4qnp_nt: 7 Other Mansleginree 6+S+9 - (67+68) Sours OATT  - 21 - Table 2.5: GROWTH OF YUGOSLAV EXPORTS BY MARKET AND COMMODITY GROUP, 1970-79 (Index Numbers of Value Series, 1970=100) Primary Manufactures Total Developed Countries 248 414 320 Developing Countries 509 592 571 Centrally Planned Economies 395 583 522 TOTAL 305 515 421 Source; Table 2.4 2.15 It is clear from these tables that over the decade exports to the developed countries have grown more slowly than those to either the LDCs or the CPEs. Since 1975 the CPEs have consistently been a slightly larger market for Yugoslav exports than the DCs. Over the period there has also been a pronounced shift toward exports of manufactures, from about 55% of total merchandise exports in 1970 to approximately 68% in 1979. 2.16 As Table 2.4 demonstrates, the commodity composition of exports differs significantly by market area. Primary product exports feature much more prominently in exports to the developed countries, while trade in manu- factures predominates in exports to the CPEs and LDCs. This has continued to be the pattern through the 1970s, and since about 1975 the CPEs have been the most important market for Yugoslav exports of manufactures. Within manufac- tures, the pattern of specialization also differs by market area. Exports of chemicals and machinery dominate in trade with the CPEs and LDCs, while the lighter manufactures are exported to the developed countries. The analysis of the earlier report showed that Yugoslavia is a relatively strong exporter of various categories of machinery and transport equipment. Even by developed country standards it is relatively specialized as an exporter of heavy electrical equipment, ships and boats, and metal products. In the lighter manufactures, Yugoslavia has been successful in furniture and wood products, certain categories of garments, and footwear. By international standards exports are extremely diversified; in general, manufactured exports have represented a spillover from production for the domestic market, rather than representing the outcome of a sustained attempt at export specialization. However, in certain sectors, notably shipbuilding, wood products and footwear, exports constitute an important part of total production.  - 22 - Performance in-Developed Country Markets 2.17 As noted above, growth in exports to the developed countries has been much slower than to the other market areas through the 1970s. This trend has been a source of concern to Yugoslav policy makers for several reasons. First, these economies offer the largest markets and are responsible for the bulk of the growth in world trade. Second, trade with the CPE market area tends to be governed by bilateral agreements which cause it to be approximately balanced while Yugoslavia has a need for uncommitted convertible currency resources to meet its debt service obligations. Finally, successful performance in industrial markets is seen as a barometer of the general competitiveness and dynamism of Yugoslav industry, and a faltering performance as connoting the opposite. 2.18 In order to develop a better understanding of the factors responsible for this slow growth, the earlier report undertook a market share analysis for selected intervals in the 1970s. In that analysis, export growth was decomposed into four components: (i) the growth of total imports into the markets being examined (the overall market effect); (ii) the deviation between the commodity distribution of Yugoslavia's exports to the markets in question and the commodity distribution of their aggregate imports (the commodity composition effect); (iii) the deviation between the markets to which Yugoslavia exports in the market area, and the overall distribution of markets in the market area (the market distribution effect); (iv) a residual, referred to here as the competitiveness effect, which incorporates the impact of all other factors such as costs, quality, government policies and the like. 2.19 The summary results of this analysis are shown in Table 2.6. The tables indicate that Yugoslav exports systematically lost market share in the markets of the developed countries between 1970 and 1977. This was partly due to an unfavorable commodity composition of exports but also reflected an adverse "competitive" or residual effect. The size of the shortfall was substantial: if Yugoslavia had retained its aggregate share in these markets, the growth in exports to them would have been 36% higher than the growtb that in fact occurred. A more detailed examination of performance in particular commodity groups in the previous report indicates that the main losses in market share arose in exports of primary products, both agricultural products and metals and minerals. This reflects several factors at work: poor output performance in several of the sectors concerned, a policy decision by the Yugoslavs to stress exports with higher domestic value added, diversion of output to the home market in response to more favorable prices and buoyant demand, and a diversion of exported output to  - 23 - more lucrative market areas, the oil exporting developing countries in the case of agricultural products, and the CPEs in the case of metals and minerals. In addition it is possible that Yugoslavia was hurt by market integration movements in Western Europe in the 1970s, including the expansion of the EEC in 1973, the reduction in tariffs between the EEC and the EFTA and the growth of preferential trading arrangements between the EEC and certain developing countries. 2.20 Performance in manufactured exports to developed countries was more satisfactory, in that growth of Yugoslav exports more than kept pace witb the growth of the market. In addition, its performance was also compared with the performance of a reference group of newly industrializing developing countries (NICs) on the same markets in the same period. On the basis of this comparison its performance was considerably weaker. This is uniformly true of all markets examined and for virtually the entire range of manufactures. While Yugoslavia may have been hurt by market access limitations in specific products and specific markets, the uniformly better performance of the NICs in all markets and virtually all categories of manufactures suggests that factors affecting the supply of exports from Yugoslavia account for much of the difference. Put somewhat differently it was a legitimate objective of Yugoslav trade policy to aim to move to exports with a higher degree of value added; such a strategy would have implied a voluntary "loss" in market share in primary products which by itself would have constituted little ground for concern. The concomitant of such a strategy however would have been a gain in market share in manufactures. The collective experience of the other NICs indicates that such a performance was a realistic possibility. A stronger performance in manufactures was also a necessity, given the balance of payments situation; however, as argued below, for reasons connected with the pattern of incentives in the economy, such a response was not forthcoming. 2.21 The general picture that emerges is one of a relatively unspecialized structure of exports representing the surplus capacity of enterprises whose primary orientation was to the domestic economy, capable of short bursts of rapid growth in response to foreign exchange shortages or unusually depressed domestic demand but sluggish over the long haul, and of an economy whose growth potential was increasingly constricted by the shortage of foreign exchange generated by this export performance. These are well-known characteristics of the export performance of an economy whose structure of incentives has been oriented toward import substitution: as has been noted, one of the self-reinforcing phenomena in such an economic regime is that the implicit discouragement of export growth tends to increase the apparent "shortage" of foreign exchange, while the evolution of import substitution means that the economy's vulnerability to shortages of imports (or foreign exchange) become progressively greater. 1/ A detailed analysis 1/ See Anne 0. Krueger, "Interactions Between Inflation and Trade Regime Objectives in Stabilization Programs" in William R. Cline and Sidney Weintraub (Eds.) Economic Stabilization in Developing Countries, Brookings Institution, Washington, D.C., 1981.  - 24 - of the nature and scale of the bias against exports is presented later in this report; the existence of sucb biases, and their consequence for the pattern of resource allocation were important underlying factors in the crisis of 1979 and go a long way toward explaining the severity of its consequences for the economy. Table 2.6: CONSTANT MARKET SHARE ANALYSIS OF YUGOSLAV EXPORTS TO DEVELOPED COUNTRY MARKETS, 1970-79 1970 1977 Actual Exports (c.i.f.) to Developed Market Economies (US$ million) 951.2 2625.6 -1970-77 Total Increase in Exports 1674.4 100.0 Due to: Overall Market Effect 2282.2 136.3 Commodity Composition Effect -345.2 -20.6 Market Distribution Effect -40.2 -2.4 Competitive Effect -222.4 -13.3 Total Sbortfall -607.8 -36.3 Source: GATT Trade System Data, equivalent to UN Trade Series D, 1979. Invisibles 2.22 Yugoslavia's balance of payments current account benefited substantially from net receipts on invisibles over the 1976-80 period. Table 2.7 below summarizes the performance of this component of the external sector. Invisible receipts during the period constituted just over half of Yugoslavia's gross current account receipts, with this share having been fairly constant over most of the 1970's.  - 25 - Table 2.7; DEVELOPMENTS IN THE INVISIBLES ACCOUNT, 1976-80 1971-75 1976 1977 1978 1979 1980 1976-80 Values (US$ millions) Receipts from Invisibles 14,662 4,410 4,769 5,775 7,732 9,194 31,880 -Nonfactor Services 7,801 2,051 2,216 2,380 3,771 4,547 14,965 Outflows on Invisibles -4,871 -1,756 -1,971 -2,714 -4,168 -5,399 -16,008 -Nonfactor Services 1/ -3,284 -888 -888 -1,051 -1,624 -1,565 -6,016 Balance on Invisibles 9,791 -2,654 2,798 3,061 3,564 3,795 15,872 Balance on Nonfactor Services 4,517 1,163 1,328 1,329 2,147 2,982 8,949 Ratios Invisibles Receipts/ Merchandise Exports 0.99 0.90 0.91 1.02 1.14 1.02 1.01 Invisibles Receipts/Total Exports 0.50 0.47 0.48 0.50 0.53 0.51 0.50 Net Invisibles/Trade Deficit 0.85 1.07 0.64 0.71 0.49 0.62 0.65 Source: Statistical Appendix Table A 3.2 2.23 The largest contribution to the invisibles account resulted from the remittances of Yugoslav migrant workers abroad. 1/ Table 2.8 presents some of the relevant statistics. In framing the plan's targets, the Yugoslav authorities had assumed that the role of workers' remittances would decline as a source of support to the trade account. Prior to the plan period, the number of workers abroad was declining at an annual rate of about 77,000 workers from its 1973 peak of 1.1 million workers. The rate of decline slowed down, however, so that by the end of the plan period it averaged about 15,000 workers a year. 2/ As a consequence while the plan had estimated the number of migrants abroad at 530,000 workers in 1980, the actual number was 780,000. The strong growth of inflows, at a measured nominal rate of about 20% per year, was sustained mainly by a revival of growth rates in GDP in 1/ These data primarily reflect movements in foreign currency deposits held in Yugoslavia, with gross increases in such deposits being regarded as an inflow and gross withdrawals being regarded as an outflow. As a number of factors can influence the role of turnovers of these deposits, it is best to regard the net movement in the stock of these deposits as the appropriate measure of remittance inflows. 2/ See Chapter VIII for a fuller discussion.  - 26 - Western Europe from the recession of 1975, altbough the various incentives offered to holders of foreign exchange deposits in Yugoslavia were also probably of importance. Apart from attractive interest rates in foreign currencies these have included privileged access to certain raw materials and equipment, and exemption from sales taxes on selected items. Table 2.8: MIGRATION AND WORKERS' REMITTANCES 1976 1977 - 1978 1979 1980 Net Worker Migrants 1/ 870 825 800 790 770 Workers' Remittances Inflows 2/ 2,269 2,430 3,240 3,773 4,450 Workers' Remittances Outflows 2/ -497 -700 -1206 -1,718 -2,548 Net Workers' Remittances 2/ 1,772 1,730 2,034 2,055 1,902 Inflows per Migrant Worker ($) 2,608 2,945 4,050 4,776 5,779 Inflows/Trade Deficit (Z) 91 55 75 52 73 Net Inflows/Trade Deficit (Z) 71 39 47 28 31 1/ Thousands 2/ In USg millions. Source: Statistical Appendix Table A 3.2 and information provided by the Yugoslav authorities. 2.24 The counterpart to gross inflows into foreign exchange deposits has been a high growth in outflows from these deposits. These outflows, measured in nominal terms, rose at an average annual rate above fifty percent in the plan period. As a result, the net offset to the trade deficit provided by net workers' remittances has rapidly decreased, from 71% in 1976 to about 31% in 1980. This factor largely accounts for the relative stagnation of the surplus on invisibles during most of the plan period. 2.25 The provision of transportation services represented another important component of the net export of invisibles. Included in this category are receipts on account of freight, passenger services, insurance, and port services. Table 2.9 provides some of the relevant statistics. During the plan period, expenditures on shipping and transport services grew a little faster than receipts on this account, partly because an increasing share of Yugoslavia's trade in goods was transported in shipping vessels of foreign registry.  - 27 - Table 2.: SHIPM4ENT AND TRANSPORTATION SERVICES 1976 - 1977- 1978 - 1979 1980 US$ millions Receipts 908 1,071 1,150 1,405 1,690 Expenditures -479 -503 '563 -674 -845 Net Receipts 49568 587 71845 Sbipping Goods Loaded 1/ 334 392 376 383 438 Goods Unloade_d 1/ 1,348 1,374 1,516 1,716 1,928 Share of World Fleet 2/ (% 0.52 0.58 0.58 0.57 0.54 Merchant Vessels Launched 2/ 583 285 242 173 123 Indices Freight Rates 3/ 105 112 118 131 145 Freight Rates 4/104 116 131 153 174 1/ Measured in thousand metric tons. 2/ Measured in thousand gross registered tons. Refers to vessels of 100 gross tons and over, excluding non-propeller and wooden vessels. 3/ Mixed currency index - 40% German marks, 60% US dollars, 1975=100 4/ US dollar-index) 1975=100. Source; Statistical Appendix Table A 3.2; UN, Monthly Bulletin of Statistics, various issues. 2.26 The tourism sector was another major source of foreign exchange; Table 2.10 records the development of this sector over the plan period. Measured in terms of the total number of nights spent by foreign tourists in Yugoslavia, there has been an average annual increase of 5.2% in the provision of such services. The number of foreign tourists, and of tourist nights, fell in 1979 as a result of the adverse impact of an earthquake in Miontenegro, but has since exceeded the highest levels that were reached in the past. The net inflow of receipts on this account has averaged about US$1 billion and, during the plan period, was a fairly dependable source of foreign exchange receipts. Official estimates of both gross and net receipts probably understate actual flows; in many resort areas, foreign currency transactions outside the official exchange market are quite common.  - 28 - Table 2.10: TOURISM SERVICES -1976 -1977 1978 1979 1980 US$ millions Receipts 802 841 1,050 1,183 1,645 Expenditures -77 -91 -120 -155 -140 Net Receipts 725 750 930 1,028 1,505 Foreign Tourists 5,572 5,621 6,385 5,966 6,412 Tourist Nights 29,368 29,026 34,866 33,482 36,984 Peak Season Hotel Occupancy Rate (%) 75.5 84.9 93.0 87.4 93.5 Source: Statistical Appendix Table A 3.2, 1980; OECD, Tourism Policy and International Tourism, various issues. 2.27 Among other components of the invisibles subaccount, Yugoslavia's provision of construction services is very important. It has not proven possible to obtain a precise measure of the export of services related to the construction of capital projects abroad, but it is estimated that net annual inflows have increased from about US$200 million in 1976 to about US$450 million in 1980. During the period, construction services were related to turnkey projects in Iraq, Cuba, Libya and the USSR, although a number of projects, especially in the construction of hotels, were undertaken in other countries. Yugoslavia has provided export credits to encourage such projects and, by 1979, had taken new initiatives in joint participation with other contractors on construction. In spite of these efforts, two major factors worked to slow the growth of such exports. First, the fairly active growth of domestic construction demand diverted interest to the home market. Second, due to this construction boom, particularly between 1977 and 1979, bottle- necks in the supply of materials, especially of pre-fabricated materials and cement, occurred not only in Yugoslavia, but also in its major suppliers. As a result, a number of unexploited opportunities existed at the end of 1980. External Debt 2.28 According to World Bank data, total medium- and long-term debt stood at a level of $5.8 billion at the end of 1975; by the end of 1980 this had risen to $15.4 billion, a increase of $9.6 billion, sixty percent higher than the amounts implicit in the plan. The average annual rate of growth in medium- and long-term debt in nominal terms was about 22% per year over the plan, the same rate as in the 1971-75 period.  - 29 - Table 2.11: EXTERNAL DEBT INDICATORS, 1976-80 1975 1976 1977 1978 1979 1980 A. Medium- and Long-Term Debt (millions of US dollars) 1. End of Period Stocks Total 5,820 7,172 8,956 11,117 13,718 15,345 Convertible currency 5,616 6,377 7,804 9,801 12,356 14,704 2. Debt-Service Payments Total 1,440 1,595 1,886 3,054 3,455 Amortization 1,138 1,228 1,308 2,237 2,379 Interest 302 367 578 817 1,076 Convertible Currency 1,186 1,492 1,774 2,287 2,899 Amortization 813 1,114 1,336 1,568 1,844 Interest 373 378 438 719 1,055 B. Short-Term Debt (millions of US dollars, convertible currency only) 1. End Period Stocks 250 350 432 868 2,050 2. Interest Payments 11 16 20 44 150 C. Memo Items (millions of US dollars) GNP 37,293 45,667 54,783 67,428 68,562 Receipts from Exports of Goods, Services and Current Transfers (Total) 9,288 10,023 11,446 14,526 18,172 Receipts from Exports of Goods, Services and Current Transfers (Convertible)l/ 6,977 8,674 9,966 11,850 14,568 D. Ratios (in percent) 1. Total MLT Debt/GNP 19.2 19.6 20.3 20.3 22.5 Total Debt/GNP 19.9 20.4 21.1 21.6 25.4 Total M+LT Debt Service/ Total Current Receipts 15.5 15.9 16.5 21.0 19.0 2. Convertible MLT Debt/GNP 17.1 17.1 17.9 18.3 21.4 Total Debt/GNP 17.8 17.9 18.7 19.6 24.4 Convertible M+LT Debt Service/ Convertible Current Receipts 17.0 17.2 17.8 19.3 19.9 1/ Estimated. Source: Statistical Appendix Tables A 4.1, A 3.1, A 3.2 and information provided by the Yugoslav authorities.  - 30 - 2.29 The ratio of nominal medium- and long-term debt to the nominal value of output, and the rate of growth of this ratio over time, while linking a stock to a flow, provides one measure of the burden of external liabilities. From Table 2.11, which presents some quantitative indicators of the debt burden, it is clear that this burden bas been growing over time. Medium- and long-term debt rose from a level of 19.2% of GNP in 1976, to 22.5% in 1980; the comparable ratios for all debt are 19.9% and 25.4%. The relative size of debt repayable in convertible currencies rose even faster. For all debt, it rose from 17.8% to 24.4%, while for medium- and long-term debt alone it rose from 17.1% to 21.4%. 2.30 Confirming this analysis of an increasing external debt burden is the ratio of debt service payments to total receipts from exports of goods and services. The M&LT debt service ratio in 1980 at 19% was about 3.5% higher than at the beginning of the Plan period, while the ratio of convertible currency debt service to convertible currency receipts had risen from an estimated 17.0% at the beginning of the plan to 19.9% at the end. 2.31 Table 2.12 provides information on the structure of borrowing by type of credit and by source over the plan. As can be seen there was a steady shift away from official sources to commercial sources of finance. Table 2.12: STRUCTURE AND TERMS OF BORROWING, 1975-80 (in percent) 1975 1976 1977 1978 1979 1980 By Type of Credit Financial Credits 36 38 35 36 37 38 Trade and Equipment Credits 62 61 63 60 59 54 Short-term Credits 2 1 2 4 4 8 By Source Official 27 27 25 23 20 20 Commercial 73 73 75 77 80 80 Average Interest Rate Publicly Guaranteed 8.1 7.4 7.3 6.8 5.8 14.6 Private, Nonguaranteed /1 8.4 7.3 7.8 10.4 12.9 15.5 Weighted Average Rate 8.3 7.3 7.6 9.3 11.0 15.2 1/ Includes variable rate debt. Source: Debtor Reporting System, World Bank, and Quarterly Bulletin, National Bank of Yugoslavia, various issues.  - 31 - There has been a fairly consistent reliance on medium and long-term financial credits which provided about 37Z of gross borrowing needs, witb the remainder being provided by trade and equipment credits of various kinds, including export credits, commodity credits and World Bank loans. The table highlights the increased use of short-term credits in recent years. This includes only the short-term external liabilities of Yugoslav banks. No information is available on short-term debt owed to suppliers. While Yugoslavia's use of short-term commercial bank credit increased in the later years of the plan, the aggregate amounts remained at acceptable levels. The buildup of short-term debt is however an indication of Yugoslavia's difficulties in obtaining medium-term financial credits adequate for its needs. 2.32 The greater reliance placed on commercial sources of funds, and the particularly rapid growth of medium and long-term obligations in the second half of 1979 and 1980 when maturities were sbortening, together with the effects of an earlier bunching of borrowing in the period following the first oil crisis, have led to a significant worsening in the maturity structure of Yugoslavia's debt. One indication of this is provided in Table 2.13, which shows that the five-year time profile ratio of Yugoslav medium- and long-term debt has increased substantially since 1971. 2.33 There were considerable shifts in the mechanisms for regulating foreign borrowing over the period, in an effort to reconcile the aims of decentralization of responsibility with the need for overall control. After 1965 enterprises were authorized to borrow abroad under the guarantee of their own business banks, without requiring the guarantee of any federal level authorities. Enterprises were, however, required to register their loans with the National Bank of Yugoslavia (NBY), to facilitate monitoring and record-keeping. Control of aggregate borrowing was regulated by the NBY by requiring borrowers to place interest-free dinar deposits with it, thereby raising the effective cost of foreign borrowing. Although this mechanism for controlling the overall level of borrowing was effective, the decentralization of borrowing to enterprises and business banks resulted in an uncoordinated approach to international financial markets particularly in solicitations for financial credits. This lack of coordination raised concern among foreign lenders on the degree of overall supervision that was in fact being provided, and occasionally resulted in Yugoslav borrowers competing against each other. With the creation of communities of interest for foreign economic relations (CIFER) in 1978 (described more fully below),  - 32 - Table 2.13: FIVE YEAR TIME PROFILE RATIO OF MEDIUM-AND LONG-TERM EXTERNAL DEBT (DISBURSED AND OUTSTANDING) (US$ millions) Beginning of Period 1971 - 1976 1981 Debt Outstanding 2,710 7,682 17,006 Projected Amortization (Five Years) 1/ 860 3,842 12,325 Five Year Time Profile Ratio 2/ 0.32 0.50 0.72 1/ Represents projected amortization of existing debt. 2/ Estimated by taking the ratio of amortization payments over the following five-years to the debt outstanding in the benchmark year. Source: World Bank, Debtor Reporting System and External Debt Tables; information reported to IMF and staff estimates. the system of compensating deposits was abolished and the allocation of borrowing rights was undertaken within the various regional CIFERs, to be consistent with the overall balance of payments position for the region within the framework of foreign exchange policy for the federation. At the same time measures were taken to improve the coordination of Yugoslav enterprises in approaching the international capital markets. These included the preregistration of all intended borrowing with the NBY (following approval by the relevant CIFER), and the joint determination by all Yugoslav banks of guidelines for minimum acceptable borrowing terms. In addition, Yugoslav banks were encouraged to approach the international markets as a consortium when soliciting financial credits on behalf of their clients, in order to improve their negotiating position. 2.34 While this remains the formal mechanism in place at present its actual workings have been shaped by the turbulent environment that has existed since its inception. The unanticipated shocks to the balance of payments in 1979 made it difficult both to adhere to the agreed framework for the year, and therefore to police performance by the republics and  - 33 - provinces. The determination to reduce the current account deficit very sbarply in the years tbereafter has made interrepublican negotiation of the distribution of borrowing rights an extremely contentious affair, leading in 1980 (and again in 1982) to a deadlock which in 1980 had to be resolved by an exceptional decision of the Federal Executive Council. It has also proved very difficult to reach medium-term agreement to govern regional borrowing rights in the 1981-85 plan. Disciplined control has been maintained over foreign borrowing even in the face of such interregional disagreements, but this has required exceptional measures rather than representing the operation of the system as intended. A further departure from the system of borrowing as envisaged has been the need for the NBY to reenter the markets as a major borrower in its own right, a development made necessary by shifts in market sentiment against lending to Yugoslavia, discussed below. This move violated a basic tenet of the new system of foreign borrowing, which was that enterprises directly, rather than state bodies should be responsible for foreign borrowing, in full recognition of the obligation this imposes on them to service the debt from their own resources. 2.35 For the reasons given above, the attitude of foreign commercial banks toward Yugoslavia has remained somewhat tentative even though Yugoslavia's borrowings from them have been in aggregate quite sizable. The evidence suggests that within the syndicated loan market Yugoslavia has throughout been a marginal borrower, liable to be rationed out in periods of market illiquidity, or when market sentiment turns against it. Thus, Yugoslavia's share of publicized Eurocurrency credits going to higher and upper middle income developing countries and centrally planned economies fell from 9.3% in 1973 to 2.6% in 1980 while its maturities and spreads have generally been worse than those of other LDCs in that category. Reasons for this status include the complexity and unfamiliarity of the Yugoslav economic system, and the perceptions of lack of coordination and control cited above. These misgivings were further buttressed by the sharp increase in the current account deficit in 1979, inability of the republics and provinces to reach agreement on the framework for policy in 1980, and the increase in inflation in that year. They were also given added force by events unfolding in Poland over the course of 1980 (and 1981) and a general reassessment of the exposure of commercial banks in Eastern Europe. As a result of these developments Yugoslavia found itself unable to borrow medium- and long-term financial credits from commercial sources in 1980 on the scale that had been planned. This experience both led to the reemergence of the NBY as a borrower, and an increased reliance on lines of credit arranged through the good offices of governments sympathetic to Yugoslavia rather than strictly on a commercial basis. 2.37 For several reasons therefore Yugoslavia's external debt position at the end of the plan was less comfortable than it had been at the beginning. While the growth in medium- and long-term debt outstanding and disbursed was no faster than in the previous plan, the shortening of maturities, the rise in international interest rates, the increasing use of  - 34 - floating rates, the slowdown in export growth and its increased orientation to the bilateral area all made for less room for maneuver, although the situation remained within manageable limits. These developments were, bowever, compounded by shifts in market sentiment against Yugoslavia, and the combination of these objective and subjective constraints on the capital account bave been a powerful determinant of the strategy of the 1981-85 plan.  - 35 - III. PLAN PERFORMANCE: THE-DOMESTIC ECONOMY A. Sectoral Growth Performance 3.01 Evidence from the 1976-80 plan period reveals that the reorientation of the structure of output (to be realized by a substantial quickening of growth rates in the priority sectors) did not occur. The data in Table 3.1 indicate that, within industry, the largest shortfalls between target and actual growth rates occurred in the priority sectors. In no priority sector, with the exceptions of oil and food processing, did attained growth reach its target level. In addition, with the exceptions of oil, chemicals and food processing, the growth rates in the priority sectors were below, not above, the average growth rate realized in industry. An even more dramatic indication of the poor performance of the priority sectors is the fact that with the exceptions of coal and food processing, actual growth rates in the 1976-80 period were significantly below the rates achieved in the 1971-75 period. 3.02 The explanations for this outcome are complex and both sector- specific and general. The designation of a particular sector as a priority sector was in part a reflection of previous difficulties encountered in developing that sector and it is not surprising that these difficulties persisted in the 1976-80 period. In the shipbuilding sector, poor output growth was directly related to the fall-off in external demand for the sector's output. In the ferrous and nonferrous metal subsectors output was capacity constrained, and the failure of output to grow was closely linked with failures in investment implementation, particularly at the primary stage. Thus iron ore production declined over the plan on account of delays in the opening of iron ore mines, and performance in the principal nonferrous ores (copper, aluminium and lead) was also extremely weak. A notable exception to this trend was the performance in coal mining where investment proceeded on target, and shortfalls in production were due to delays in the completion of associated power plants rather than any weakness in the coal sector itself. Overall it seems fair to say that output difficulties reflected difficulties in investment choice and implementation; the sources of these difficulties are discussed later. 3.03 Performance in the nonindustrial priority sectors of agriculture, transportation and tourism was also below target; in the case of agriculture growth in the 1976-80 period was significantly below the rate achieved in the 1971-75 period. As discussed in greater detail in Chapter VII below, the major shortfall was in wheat and rye, although growth in livestock production was also lower than in 1971-75. The wheat shortfall was common to both the social and individual sectors, witb social sector wheat production declining by about 31% between 1976 and 1979. The poor performance in wheat reflected adverse climatic factors but also a low relative price for wheat which induced a shift in cropping patterns, and low fertilizer usage in the individual sector.  - 36 - Table 3.1: ACTUAL AND TARGET GROWTH RATES BY SECTOR, 1971-1980 1/ 1 2 3 4(-3-2) 1971-75 1976-80 1976-80 -1976-80 Actual Planned Estimated Estimated-Planned Priority Sectors Industrial Electrical Energy 8.5 10.0 8.0 -2.0 Coal 4.0 9.5 5.3 -4.2 Oil 7.7 6.0 6.5 .5 Ferrous Metallurgy 8.8 11.0 4.6 -6.4 Nonferrous Metallurgy 7.4 11.0 3.6 -7.4 Nonmetallic Minerals 7.3 9.0 5.6 -3.4 Shipbuilding 4.6 9.0 -5.6 -14.6 Chemicals 12.0 14.0 9.7 -4.3 Food Processing 6.7 7.3 7.2 -.1 Nonindustrial Agriculture 2.9 4.0 1.9 -2.1 Transportation 5.7 7.5 5.3 -2.2 Trade & Tourism 5.7 6.7 5.6 -1.1 Nonpriority Sectors Industrial Metal Industry 8.8 8.5 8.0 -.5 Electrical Equipment 10.9 9.0 8.5 -.5 Construction Materials 8.8 9.0 6.8 -2.2 Wood 7.3 6.0 6.5 .5 Paper 6.6 6.0 7.8 1.8 Textiles 7.2 4.0 4.7 .7 Leather 7.6 4.0 4.2 .2 Rubber 9.6 4.0 10.8 6.8 Printing 5.1 4.0 2.3 -1.7 Tobacco 5.7 5.0 7.4 2.4 Misc. Manufacturing 9.7 4.0 8.0 4.0 Nonindustrial Forestry 2.7 2.0 2.0 0.0 Construction 4.0 7.5 7.9 *4 Crafts 7.7 7.2 5.2 -2.0 Public Utilities 8.8 7.5 5.9 -1.6 Total Industry 8.1 8.0 6.7 -1.3 Total Social Product 6.5 7.5 6.4 -1.1 I/ Industrial growth rates are growth rates of the gross physical volume of production. Growth rates of nonindustrial sectors are growth rates of real gross social product, 1975 prices. Source: Data provided by the Yugoslav authorities.  - 37 - Table 3.2: GROWTH OF RAW MATERIAL AND PROCESSING SUBSECTORS IN PRIORITY INDUSTRIAL SECTORS (Physical volume of production, in percent) 1971-75 1976-80 /1 Coal 4.0 5.3 Extraction 3.5 4.2 Processing 8.1 16.4 Oil and Gas 7.7 6.5 Extraction 6.0 3.0 Refining 8.0 8.3 Ferrous Metallurgy 8.8 4.6 Iron ore mining 7.0 -3.4 Iron & steel industry 8.9 5.2 Nonferrous Metallurgy 7.4 3.6 Ore mining 4.9 3.3 Basic industries 11.0 3.5 Smelting, alloying and refining 6.0 5.3 Nonmetallic minerals 7.3 5.6 Extraction 4.5 -1.3 Manufacture 8.0 7.0 Chemicals 12.0 9.7 Manufacture of basic chemicals 8.2 8.0 Processing 13.7 10.0 All raw material sectors 6.6 6.0 All investment goods sectors 9.0 7.4 All consumer goods sectors 7.8 6.3 All industry 7.7 6.4 /1 Estimated. Source: Statistical Yearbook of Yugoslavia, various issues and Indeks.  - 38 - 3.04 Given the importance placed on import substitution in intermediate inputs and energy in the 1976-80 plan, it is useful to examine the growth rates realized in the energy and raw materials sectors in somewhat more detail. Table 3.2 presents actual growth rates for the 1971-75 period and the 1976-80 period for six intermediate and energy sectors designated as priority sectors in the 1976-80 period. Each intermediate sector in turn is subdivided into two parts; that part of the industry involved in the provision of the basic raw material input and that part of the industry involved in the processing of the input or in the manufacture of products embodying it. Thus, for example, the coal sector is subdivided into coal production and coal processing and the chemicals sector into basic chemicals production and chemical processing. 3.05 In each of these sectors, comparison of growth rates in the component subsectors during each of the two periods reveals a persistent pattern for growth at the processing stage of the industry to outpace growth in the domestic raw material input base. As a consequence of these imbalances, assuming a fairly stable relationship between processed output and raw material input, the dependence of these sectors on imported inputs of basic raw materials must have remained substantial. The extent of this dependence is suggested by the figures in Table 3.3 which indicate that the direct intermediate import requirements in these sectors have consistently been greater than the direct intermediate import requirements for all productive sectors. In each of these sectors, intermediate imports are mainly imports of raw materials for further processing by domestic producers. Table 3.3; RATIO OF DIRECT IMPORTS OF INTERMEDIATE INPUTS TO GROSS DOMESTIC PRODUCTION (Current price value shares, in percent) 1966 1968 1970 1972 1974 1976 Coal 12.3 12.2 11.6 15.4 12.9 18.1 Oil and gas 25.9 26.8 29.4 28.8 43.3 43.1 Ferrous Metallurgy 9.5 8.8 17.0 17.5 23.7 11.8 Nonferrous Metallurgy 9.5 12.3 25.4 26.4 24.3 11.5 Nonmetallic Minerals 4.3 5.9 7.0 9.4 13.1 7.9 Chemicals 17.5 n.a. 23.9 25.0 29.8 26.6 Total productive sectors 6.3 6.6 8.5 9.5 11.8 7.8 Source: Mate Babic, Strukturna Analiza Kretania Vanisko-trgovinske Razmjene Jugoslavije, Ekonomaki Institut, Zagreb, 1979 (Structural Analysis of the Movements of the Yugoslav Foreign Trade, Economic Institute, Zagreb, 1979).  - 39 - 3.06 The imbalances between growth rates in raw materials and growth rates in processing evident in the priority industrial sectors were also evident in the overall behavior of the economy over the 1976-80 period. On an aggregate level, as the data in Table 3.2 indicate, the growth of output in all sectors producing intermediate and energy products fell from the 6.6% realized in 1971-75 to 6.0% in 1976-80. As in the earlier period, this was below both the average growth rate for all industry and the growth rates of sectors producing investment and consumer goods. B. Investment 3.07 As mentioned before, the main tool employed by the Yugoslavs for restructuring the economy in the 1976-80 plan was the priority allocation of investment resources. Under the new institutional arrangements set in place by the 1974-76 reforms, banks, governments and enterprises were obliged to formulate and execute a number of self-management agreements and social compacts to set aside the necessary investment resources for priority sector development. By the design of these new arrangements, investment projects for priority sectors were to be prepared first and their financing was to be mancatory while financing for nonpriority investment projects was to depend on what was left over. The evidence from the 1976-80 period indicates that there was some slippage between system design and system practice. 3.08 The data in Table 1.3 indicate that during the 1976-80 period, the new institutional mechanism succeeded in raising the share of the priority sectors in economic investment in fixed assets by 6.2 percentage points relative to their 1971-75 share. 1/ This increase is substantial, although it was only about 55% of the planned increase of 11.3 percentage points. Nonetheless, the aggregate evidence supports the conclusion that the new investment allocation mechanism was successful in directing a significantly larger portion of the total investment effort to the priority sectors. 3.09 Behind this success at the aggregate level, however, lie important divergences between planned and actual investment at the sectoral level as the data in Table 1.3 indicate. In the priority industrial sectors -- electrical energy, coal, oil and gas, ferrous metals, nonferrous metals, nonmetallic minerals and basic chemicals -- realized investment shares apparently increased as planned, although the actual increase tended to overshoot the plan target (as in electrical energy), or to undershoot the plan target (as in oil and gas). Overall, the share of these basic industries in total economic investment in fixed assets is estimated at 33.5% over the period compared to a target share of 32.9%. By contrast, investment in machinery and shipbuilding, agroindustry, highway transportation and tourism moved counter to plan targets. The investment share of machinery and shipbuilding 1/ Throughout this section, investment refers to investment expenditures, rather than realized investment. As is discussed later, this distinction is important given the long and growing lags in project completion that characterized the 1976-80 plan.  - 40 - rose by 0.3 percentage points instead of declining by 1.2 percentage points as planned; the investment share of agroindustry, slated to increase by 1.9 percentage points, declined slightly; and the investment shares of highway transportation and tourism, targeted to remain roughly at their 1971-75 levels, in fact fell. 3.10 It should be emphasized that these findings are based on provisional data through 1980. Available evidence on the pattern of investment through 1979 suggests greater divergence between actual and planned investment shares, at least within the industrial sector. The discussion on the industrial sector in Chapter VI indicates, for example, that through 1979, the share of the energy and raw material sectors in total industrial investment rose from its 1971-75 level of 45.7 percent to a 1976-79 level of 49.0 percent, which was still sharply below the 1976-80 target level of 63.3 percent. This evidence, when contrasted with the preliminary evidence through 1980 contained in Table 1.3, suggests that there may have been a sustained investment drive in favor of basic and raw material industries in the last two years of the 1976-80 plan period. If the preliminary figures through 1980 are reliable, then this drive succeeded in raising the share of these industries in total economic investment to its target level by the end of the plan period. Nonetheless, even according to these figures, the share of these industries in total industrial investment must have fallen short of the targeted share because the share of the nonpriority industrial sectors did not fall as sharply as planned over the entire period. 3.11 Taken together, the evidence discussed here suggests three tentative conclusions: first, over the entire 1976-80 plan period, the target investment share in total productive investment of the basic energy and raw material sectors taken as a group was realized, although there were divergences between actual and target shares for individual sectors; second, the actual investment share for other priority sectors taken as a group fell short of target, with the result that the actual investment share for nonpriority sectors, mainly within industry, exceeded target; and third, as a result of the preceding two trends, the share of basic industries in total industrial investment, although rising as planned, fell short of its target share because of the greater than planned investment in nonpriority industries. 3.12 The structural redirection of the investment effort was acbieved within the context of an increase in the overall fixed investment effort relative to national output, as the data in Table 3.4 indicate. According to three different measures of national output -- gross domestic product, gross national product, and gross material product -- investment in fixed assets absorbed a higher fraction of resources than had been the case in the  - 41 - 1971-75 period. In terms of the Yugoslav gross material product measure, the share of gross investment in fixed assets rose from 29.7% during the 1971-75 period to an estimated 33.7% over the 1976-80 period, nearly 3 full percentage points above the 1976-80 plan target of 31.0%. As is discussed below, failure to control the overall investment effort via the capital allocation mechanism had important implications for macroeconomic stability and balance of payments performance. Table 3.4; IhDICAIORS OF THE OVERALL INVESTMENT EFFORT, 1971-80 (in percent) 1971-75 1976 1977 1978 1989 1980 1/ 1976-80 /1 (Actual) (EstimatedY Gross investment in fixed and working capital as a percentage of: GDP (1972 prices) /2 30.0 31.4 34.3 30.9 32.2 31.3 32.0 GNP (1972 prices) /2 28.3 29.5 32.0 28.8 30.1 28.4 29.8 GMP (1975 prices) /2 39.3 37.7 41.7 38.8 41.3 40.2 40.0 Gross investment in fixed assets as a percentage of: GDP (1972 prices) /2 27.9 30.2 30.7 31.6 29.8 29.3 30.3 GNP (1972 prices) /2 26.3 30.9 31.5 32.3 30.6 29.5 28.1 GMP (1975 prices) /3 29.7 32.3 33.4 35.5 34.7 33.1 33.7 /1 Preliminary estimates for 1980. /2 World Bank calculations of GDP, GNP and fixed investment. /3 Information provided by Yugoslav authorities. 3.13 In keeping with past trends, during the 1976-80 period, nearly 40% of this fixed investment effort was directed into noneconomic investment, of which about 76% went into housing and related public utilities. As the data in Table 3.5 indicate, this performance was broadly in line with plan tar- gets, altbougb the sbare of economic investment was slightly above its plannea sbare, and the sbare of bousing and utilities in total noneconomic investment was 4.3 percentage points below the planned sbare. The sbortfall in housing performance was attributable to the larger than planned sbare of the social sector, mainly government bodies, in total nonproductive invest- ment and a resulting boom in government office building and infrastructure (including defense). The continuing large sbare of noneconomic investment in total fixed investment provides the backdrop against which the priority allocation of investment must be understood. Once noneconomic investment is added to the picture the sbare of the priority sectors comes to 36% of  - 42 - Table 3.5; FUNCTIONAL DISTRIBUTION OF REAL INVESTMENT EXPENDITURES, 1971-80 (in percent; underlying data in 1975 prices) 1971-75 1976-80 1976-80 1976 1977 1978 1979 1980 Actual Actual Planned Total investment 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Fixed capital 75.5 84.5 79.8 85.6 80.1 91.5 84.0 82.3 Working capital 24.5 15.5 20.2 14.4 19.9 8.5 16.0 17.7 Total fixed capital formation 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Economic 60.1 60.7 61.2 58.6 61.5 61.5 61.3 60.8 Noneconomic 39.9 39.3 38.8 41.4 38.5 38.5 38.7 39.2 Total noneconomic fixed capital formation 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Social 59.7 67.7 58.7 65.9 65.9 68.0 69.2 65.2 Private 40.3 32.3 41.3 34.1 34.1 32.0 30.3 34.8 Share of Housing and utilities in noneconomic fixed capital formation 78.1 75.6 79.9 77.0 76.8 73.9 74.9 76.0 Source; Information provided by the Yugoslav authorities. total fixed investment for the 1976-80 period and the share of the nonpriority sectors in total fixed investment is only 25%. Both of these shares are less than the share of overall noneconomic investment in total fixed investment. Furthermore, the share of noneconomic, nonbousing investment in total fixed investment implied by the figures in Table 3.5 is 9.6% of total fixed investment, and this alone is larger than the share of any single priority sector in total fixed investment. The only priority sector that comes close to this investment share is electricity with a share in total fixed investment of 9 percent during the 1976-80 period. C. Capital Efficiency 3.14 With these observations in mind, the economy-wide and sectoral ICOR calculations in Table 3.6 can be used as a rough guide to changes in the efficiency of capital use over the 1970s. The limitations of this measure are numerous and well-known, particularly in a period with sharply varying rates of capacity utilization. In addition over a period as long as a decade shifts in the output mix of particular sectors, and in their technology, can also have an impact. Accordingly, only gross trends should be accorded  - 43 - significance. At the economywide level, the numbers indicate a small increase in the ICOR for productive investment in the 1976-80 period as compared to the 1971-75 period. In contrast, as the ICOR for total investment indicates a small decline between the two periods. Taken together, the two economywide estimates seem to imply that the efficiency of capital showed some gain in nonproductive uses and some deterioration in productive uses over the time period considered, but differences in the data coverage rule out any definite conclusion. 3.15 When using estimates of incremental capital output ratios as approximate measures of efficiency of capital use in Yugoslavia, it is important to keep in mind the large share of noneconomic investment in total fixed investment during the 1976-80 period. The output of housing services and governmental services (in education, health, culture and defense) that are the result of sustained noneconomic investment are notoriously difficult to estimate in any system. Indeed, as noted earlier, the Yugoslavs do not include these outputs in their gross material product accounts, so that their output estimates ignore the substantial value of service outputs created by nonproductive investment. Consequently, when looking at economy-wide performance, ICOR calculations using estimated GDP accounts prepared by the World Bank are preferable, although even these may underestimate the imputed value of housing services. When looking at the performance of economic investment in individual sectors, however, the Yugoslav data are the only data available for such an examination. 3.16 The sectoral ICORs show a sharp deterioration in the efficiency of capital use in most of the priority sectors in the 1976-80 period, when compared to the 1971-75 period. The only exceptions are the coal mining and processing sector and the machinery and shipbuilding sector, which together accounted for about 10% of total priority investment in the plan. The deterioration is especially severe in oil and gas, nonferrous metals, basic chemicals, nonmetallic minerals sector and in agroindustry (agriculture, food processing, beverages and tobacco). The results are not surprising in light of the poor output performance of most of the priority sectors (see Table 3.1) and the substantial allocation of investment expenditures to them. Nonetheless, the results for most of the priority sectors are quite dramatic in themselves, especially since the overall ICOR for productive investment increased only slightly during the 1976-80 period. The most striking increase is in the nonferrous metallurgy sector, where the ICOR increased sixfold; however, the ICOR more than doubled in the oil and gas sector and in the nonmetallic mineral sector as well. Of greater quantitative significance perhaps is the 50% increase in the ICOR of the electric energy sector, which accounted for 25% of priority investment. The very sharp increases in the ICORs of most of the priority sectors (which together accounted for nearly 60 percent of all productive investment) imply substantial reductions in the ICORs of nonpriority users of investment funds. 3.17 The perplexing question is why the efficiency of capital, as approximated by these ICOR calculations, seems to have deteriorated to such an extent in the priority sectors over this period, and why this  - 44 - Table 3.6: INCREMENTAL CAPITAL-OUTPUT RATIOS, 1971-1980 1971-75 1/ 1976-80 2/ (actual) (estimate) A. Economy-wide ICORs 1. Economic investment only 3/ 4.18 4.34 2. Total Investment 4/ 5.84 5.69 1976-80 as 1971-75 1/ 1976-80 2/ a ratio of (Actual) (Estimate) 1971-75 B. Sectoral ICORs (Productive Sectors Only) 5/ Electrical energy 13.34 19.87 1.49 Coal 9.27(9.60) 6/ 9.09(9.57) .98(.99) Oil & gas 4.42 11.85 2.68 Ferrous Metallurgy 5.76 7.13 1.24 NonferrousMetallurgy 5.78 36.86 6.38 basic chemicals 1.30(3.63) 7/ 2.27(7.13) 1.75(1.96) Nonmetallic minerals 1.24 3.19 2.57 Machinery & shipbuilding 1.65 1.80 1.09 Agro-industry 3.25 5.78 1.78 highway transportation /8 n.a. n.a. n.a. Foreign tourism /8 n.a. n.a. n.a. All productive sectors 4.18 4.34 1.04 1/ Change in ouput ( A Q) and change in capital stock ( A K) are both measured over the 1971-75 period. 2/ A Q and A K are both measured over the 1976-80 period. The 1980 figures are preliminary. 3/ Gross investment in fixed assets as a ratio of the change in gross material product (1975 prices, Yugoslav estimates). 4/ Gross investment in fixed assets as a ratio of the change in gross domestic product (1972 prices, World Bank estimates).  - 45 - Table 3.6 (continued) 5/ A K for eacb sector was estimated by taking that sector's sbare in total productive investment in fixed assets over the relevant time period. A Q for each sector was estimated by taking that sector's cbange in gross material product over the relevant time period. Since the gross material product accounts were reported in 1972 prices (Statistical Yearbook 1981, Table 107-1), while the investment series were reported in 1975 prices (Yugoslav sources), it was necessary to convert the material product estimates for each sector to 1975 prices, using the producer price index for that sector. 6/ The figures in parentheses are for coal mining only wbile the otber figures are for botb coal mining and processing. It appears from information on the use of investment resources in coal that the bulk of investment resources directed to the coal industry went to coal mining. 7/ The figures in parentbeses are for the production of basic chemicals only while the otber figures are for botb the production and the processing of basic chemicals. It appears from information on the use of investment resources in cbemicals that the bulk of investment resources directed to the cbemical industry went to the production of basic chemicals. 8/ Because data on gross material product for bigbway transportation and foreign tourism are not available separately, these calculations could not be made. Sources: Statistical Yearbook of Yugoslavia, 1981; World Bank estimates and information provided by the Yugoslav autborities.  - 46 - deterioration was specific to the priority sectors rather than to all productive sector investment. A variety of factors suggest themselves. First, as a result of nearly guaranteed access to investment funding, certain priority sector projects appear to have been poorly designed, initiated before supporting technical and market studies were completed, and haphazardly implemented. Second, the sheer scale of the priority projects was often larger, making for more complex management requirements and greater risks of slippage. Third, the majority of the ore bodies involved in the priority extractive metal and mineral industries were in the less developed regions (LDR) with their weaker physical and institutional infrastructure, altbough a significant exception to this was coal mining which, though also predominantly located in the LDR, apparently expanded its facilities without difficulty. Fourth, there appears to have been growing competition for investment goods and finance among investors in both the priority and nonpriority sectors once macroeconomic policy became expansive in 1977 and the investment boom began in earnest. This competition expressed itself in inflation, cost overruns and delay. It seems that the larger priority sector projects were particularly vulnerable to these cost overruns, estimated in aggregate at one-third or more of originally programmed dinar costs. The financial allocation mechanisms failed to allocate the amounts needed to permit priority projects to be completed; instead, resources continued to be stretched too thinly among a large number of projects. Thus, while in theory the brunt of adjustment was expected to fall on nonpriority investment projects, the rise in ICORs in all priority sectors as well as other evidence suggests that much of the burden of adjustment fell largely on them. 3.18 It is likely therefore that rising ICORs in the priority sectors, (calculated as these are on the basis of investment expenditure and not installed capacity), primarily reflect increased gestation lags and a large volume of incomplete investment in these sectors at the end of the plan, although as documented in Chapters VI and VII, there is also evidence of poor project choice in several of the priority sectors. This general finding (which accords with most Yugoslav diagnoses of performance in the 1976-80 plan) is substantiated by evidence from particular sectors. For example, in the case of electric power generation only 50% of the hydro capacity which was planned to be commissioned was actually commissioned, and the average delay per plant was of the order of 1.8 years. An analysis of the causes of these delays indicates that the two principal reasons for the slippages in commissioning dates were delays in the delivery of major items of equipment and shortages of finance. In turn, the most frequently quoted reason for the delays in delivering equipment was that domestic suppliers were overburdened by too many orders. On the financial front substantial price increases on both foreign and domestic equipment increased financial requirements, which could not be easily made up from either domestic and foreign sources. The story is similar in the case of thermal plants. Thus, shortages of both domestic and foreign credit delayed commissioning of a 50 MW district heating station in Ljubljana; two lignite-fired plants due to be commissioned in the 1976-80 period in Macedonia had to be slipped into the 1981-85 plan on account of delays in delivery of equipment, civil works problems (resulting from the large construction program) which caused shortages of essential  - 47 - equipment and manpower, and financial difficulties. Though all of the planned foreign credits for the construction of these plants were realized, many of the expected local credits were late, reflecting the general shortage of funds for investment projects. 3.19 Similar examples of cost overruns, both reflecting and leading to implementation delays and cash flow problems, can be found in a variety of sectors and are well known witbin Yugoslavia. The general conclusion supports the Yugoslav view of the matter: investment was begun on too broad a front to be sustainable, and the consequence was delay, inflation, balance of payments difficulties and a decline in investment efficiency. The policy framework that contributed to this outcome is discussed in detail in Chapter IV. D. Saving Performance Aggregate Trends 3.20 The growing investment effort of the 1976-80 years was sustained by substantial domestic and foreign saving. The resulting overall increase in the national saving rate as a percentage of gross material product, as well as the rising share of foreign saving in the financing of investment are evident in the figures in Table 3.7. 1/ According to these figures, gross investment rose from an average of 39.0% of gross material product between 1971 and 1975 to an average of 43.0% between 1976 and 1979. In total investment financing, the share of domestically generated saving (including saving out of remittances by households) fell from an average of about 96% between 1971 and 1975 to an average 92% between 1976 and 1979, while the share of foreign borrowing doubled from an average of about 4% to an average of about 8%. The sharp increase in foreign borrowing filled a gap generated by the inability of domestic saving to keep pace with the investment drive. Consistent with this failure, the shares of the two largest domestic contributors to gross savings -- social sector enterprises in the economic sector and households -- both declined from their 1971-75 averages. Unfortunately, the flow-of-funds data on which these conclusions are based are only available through 1979. As a result of both the sharp decline in foreign capital inflows and the apparent increase in enterprise saving rates due to compulsory income controls, 2/ the share of the foreign sector in gross investment finance must have declined while the share of the social sector enterprises must have increased in 1980. Thus the entire 1976-80 period probably does not differ from the 1971-75 period as much as the 1976-79 figures in Table 3.7 suggest. 1/ These data are derived from flow-of-fund accounts rather than national accounts because of the interest in analyzing saving by sector. As noted in Schrenk, et. al. (p. 137), savings rates measured in this way do not agree with savings rates measured from the national accounts. However, as the data in Table 3.7 are internally consistent, they should provide an accurate reflecton of trends over time. 2/ See Chapter VIII.  - 48 - Table 3.7: INDICATORS OF AGGREGATE SAVING PERFORMANCE, 1971-79 (Current prices, billions of dinars) 1971-75 1976-79 1976 1977 1978 1979 Total gross national saving 624.6 1357.8 246.5 282.5 369.4 549.8 of which social sector 422.1 916.1 166.5 194.0 241.3 314.3 private sector (households) 202.5 441.7 80.0 88.1 128.1 145.5 Foreign borrowing 25.2 111.3 -3.0 32.3 16.3 65.7 Gross investment 1/ 649.0 1469.0 243.6 314.4 385.7 525.3 Distribution of Saving (X) Economic Enterprises 51.5 49.1 48.1 48.1 50.8 49.4 Other 13.4 12.8 20.2 13.6 11.8 10.3 Households 31.2 32.5 32.9 28.0 33.2 27.8 Foreign Borrowing 3.9 7.6 -1.2 10.3 4.2 12.5 Gross national saving 1/ as a percentage of gross material product 37.5 40.0 41.6 38.4 40.9 39.5 Gross investment as a percentage of gross material product 39.0 43.0 41.1 42.8 42.8 45.1 1/ As measured from flow-of-funds. Sources: Flow-of-Funds tables, National Bank of Yugoslavia, Quarterly Bulletin, October 1980 and Schrenk et. al., pp. 137-138. Household Savings 3.21 Domestic saving performance over the 1976-80 period can be better understood by examining the saving behavior of households and enterprises in somewhat greater detail. The share of households within gross national saving remained above its 1971-75 average at 32.5% during the 1976-79 period. The evidence on household saving rates is somewhat contradictory and difficult to interpret. The flow-of-funds data on which the gross saving rates are based indicate that gross bousehold saving as a percentage of gross household receipts declined from 18.3% during the 1971-75 period to 17.3% during the 1976-79 period. These figures suggest a drop in the saving effort of households during the later period. In contrast, the Yugoslav accounts on household income and saving reveal an increase in the saving effort of house- holds relative to the 1971-75 period, as the data in Table 3.8 indicate. As  - 49 - the household accounts provide a better conceptual basis on which to assess household saving behavior than the flow-of-funds accounts, it seems reasonable to conclude that the household saving effort probably increased somewhat in the second half of the 1970s. It is also interesting to note that the household saving rate appeared to peak in 1978 and to begin a decline back to its 1976 level by 1980. The timing of the peak and its gradual decline coincides with the peak in the growth of real disposable household receipts in 1978 and its decline to a negative rate of growth in 1980. Enterprise Savings 3.22 As in the past, so during the 1976-80 period, Yugoslav policymakers continued to criticize social sector enterprises in the Table 3.8: HOUSEhOLD SAVING RATES (1971-80) (Current price estimates, in percent) 1971-75 1976-80 1976 1977 1978 1979 1980 Gross household saving as a share of gross household receipts /1 18.3 17.3 17.7 16.1 18.7 16.6 n.a. Gross household invest- ment as a share of gross household saving /1 59.1 56.8 60.0 58.3 51.8 51.7 n.a. Household saving (including construction expenditures) as a share of disposable household income /2 15.5 17.6 16.4 17.0 20.0 17.6 16.8 (15.4)/3 (17.3)/3 Household saving (excluaing construction expenditures) as a share of disposable household income /2 8.2 10.6 9.4 10.1 13.3 10.6 9.8 /1 Derived from flow-of-funds accounts. /2 Disposable household income is calculated as total household receipts minus net consumer credit minus tax payments. Derived from balance of incomes and expenditures of households. /3 Household income is calculated as total household receipts minus tax payments but including net consumer credit. Source: National Bank of Yugoslavia flow of funds data, and information provided by the Yugoslav authorities.  - 50 - Table 3.9: SAVING BEHAVIOR OF SOCIAL SECTOR ENTERPRISES, 1971-80 (Current price estimates, in percent) 1971-75 1976-80 1976 1977 1978 1979 1980 World Bank Concepts 1/ Gross Enterprise Saving as a sbare of disposable enterprise income 42.7 45.1 41.4 42.4 44.0 46.4 50.8 Net enterprise savings as a share of disposable enterprise income 24.3 26.8 20.9 24.1 26.3 28.9 34.0 Collective consumption funds as a share of net enterprise saving n.a 42.4 40.1 41.5 46.5 45.0 38.7 Yugoslav Concepts 1/ Accumulation as a share of dohodak n.a. 12.0 10.3 11.3 10.6 11.9 15.8 Accumulation as a share of cist dohodak n.a 15.9 12.4 14.1 14.6 16.6 22.0 1/ Please see text for definitions. Source: Schrenk, et. al., Table A.9; Statistical Yearbook of Yugoslavia, 1981; information provided by the Yugoslav authorities. productive sectors for low savings rates that contributed to the inadequate volume of domestically generated saving. Consistent with this view, the share of productive sector enterprises' savings in the gross financing of investment fell from an average of 51.5% in the 1971-75 period to 49.1% in the 1976-79 period according to the flow-of-funds data in Table 3.7. Also, according to this source, the share of enterprise saving in total domestically generated saving fell from 53.6% to 52.7% over the same two periods. 3.23 Once again, the best way to evaluate the enterprise saving effort is to use enterprise income accounts rather than the flow-of-funds accounts, and this is done in Table 3.9. Information drawn from the former source reveals somewhat different tendencies depending on the definitions of enterprise saving and income used in calculations of the enterprise saving rate. Following World Bank definitions,l/ gross enterprise saving is defined 1/ See Schrenk, et al., Table A.9  - 51 - as the sum of depreciation (whether compulsory or voluntary) and enterprise allocations to funds (including the collective consumption fund), while enterprise disposable income is defined as gross saving plus net personal incomes paid to workers. On these definitions, the data indicate that the gross enterprise saving rate increased from an average rate of 42.7% ovLr the 1971-75 period to an average of 43.6% over the 1976-79 period and to an estimated 45.1% over the 1976-80 period. The larger increase over the 1976-80 period reflects the imposition of effective incomes policies in 1980. If enterprise saving is defined net of all depreciation (including voluntary depreciation, which during the 1976-80 period was an estimated 4.7% of total depreciation), then the enterprise saving rate as a percentage of disposable enterprise income also rose from about 24.3% during the 1971-75 period to 25.1% over the 1976-79 period or 26.8% over the 1976-80 period, the latter increase again reflecting the 1980 income controls. Data on the composition of net enterprise saving during the later periods suggests that an increasing portion of this saving was set aside to finance enterprise collective consumption investment expenditures rather than enterprise expenditures on productive capital. 3.24 The Yugoslav accounting conventions applied to the measurement of enterprise saving and enterprise income are different from the ones used by the World Bank. According to World Bank conventions, disposable enterprise income is defined exclusive of all fiscal and parafiscal contributions (including all taxes and interest and principal repayments on loans) but inclusive of depreciation, reflecting the view that depreciation funds are available to the firm to finance its investment. In contrast, the Yugoslavs use two income concepts, both of which differ from the World Bank concept. Enterprise income (dohodak) is measured by gross value added minus compulsory depreciation. This measure of income includes all fiscal and parafiscal contributions that would normally be excluded in the calculation of enterprise income in a conventional profit and loss statement and that are excluded in the World Bank concept. The 1World Bank concept is much closer to what the Yugoslavs call net income (cist dohodak) which is the Yugoslav measure of income minus these charges and voluntary depreciation. This net income concept is identical to the World Bank disposable income concept except that the latter includes depreciation whereas the former does not. Finally, the Yugoslavs measure saving, called accumulation, as voluntary depreciation plus allocation to funds excluding the collective consumption fund. Using the Yugoslav concepts, the available data reveal that the accumulation rate out of income was relatively constant at about 11% during the 1976-79 period and then rose to an estimated 15.8% in 1980 as a result of the controls on personal incomes. During the same period, the accumulation rate out of net (cist) income rose gradually from 12.4% in 1976 to 16.6% in 1979, and then shot up to an estimated 22% in 1980, also as a result of controls. These saving rate calculations, using the Yugoslav concepts support the general conclusion based on World Bank concepts -- namely, that enterprise saving rates rose gradually over the 1976-79 period, increasing sharply in 1980 as a result of income controls. Unfortunately, comparable data on saving rates calculated according to the Yugoslav concepts are not available for the 1971-75 period, so it is impossible to determine whether the 1976-80 rates rose or fell relative to their 1971-75 levels.  - 52 - 3.25 Summarizing the above discussion, it seems reasonable to conclude that both gross and net enterprise saving rates out of income remaining to enterprises, after payment of material costs and fiscal and parafiscal contributions, tended to increase slightly over the 1976-79 period, jumping sharply in 1980 as a result of income controls. This conclusion can however also be reconciled with the findings reported in Chapter VIII below, which indicate that the enterprise savings effort, measured as gross accumulation (compulsory depreciation plus allocations to funds) as a ratio of enterprise value added, actually fell in the 1976-78 period relative to the 1971-75 period. Using figures that cover only social sector enterprises in economic activities and examining the distribution of enterprise value added over these time periods in Table 3.10, it appears that gross saving fell from about 26% of value added in the 1971-75 period to about 22% over the 1976-78 period and to about 23% over the 1976-79 period. Examining the other elements of value added in Table 3.10, however, it is evident that this decline in the gross enterprise saving rate was not matched by any noticeable increase in the share of net personal incomes in total value added but rather by an increase in the share of fiscal and parafiscal contributions in the total. These results, along with the results on enterprise saving rates from lable 3.9, indicate that the reported drop in the gross accumulation rate as a share of value added was not the result of a decline in the rate of enterprise saving out of disposable income (remaining after its payment of fiscal and parafiscal contributions) but the result of an increasing burden of such contributions on total value added. Table 3.10: DISTRIBUTION OF VALUE-ADDED OF SOCIAL SECTOR ENTERPRISES IN THE PRODUCTIVE SECTOR 1/ 1966-70 1971-75 1976-79 1976 1977 1978 1979 Social Product 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Depreciation 10.4 11.3 10.8 11.9 10.7 10.5 10.2 Net Social Product 89.6 88.7 89.2 88.1 89.3 89.5 89.8 Fiscal and parafiscal 39.6 39.0 41.5 42.1 40.7 41.5 41.7 contributions Net personal incomes 36.2 34.7 35.1 36.1 36.2 34.5 33.6 Allocations to funds 14.0 15.0 12.5 9.9 12.4 13.4 14.4 Gross saving 24.4 26.3 23.3 21.8 23.1 23.9 24.6 /1 Figures do not add up to 100 in all cases due to rounding errors. Source: 1966-75 figures, Schrenk, et al., Table 7.8; 1976-78 figures provided by Yugoslav authorities; 1979 figures, Statistical Yearbook, 1981, Table 113-4.  - 53 - 3.26 Despite what appears to bave been a slight increase in enterprise saving rates out of disposable enterprise income over the 1976-79 period, the self-financing rate of social sector enterprises -- measured as gross saving compared to gross enterprise investment in fixed and working assets -- dropped from a 1971-75 rate of 74.0% to a 1976-79 rate of 69.6%. This drop ran counter to one of the main objectives of the 1974-76 reforms: to increase the self-financing of investment through the pooling of enterprise resources. 3.27 The inadequacy of enterprise saving cited by Yugoslav policymakers is more apparent if one compares enterprise saving to the value of the capital stock in the enterprise sector. Some data relevant to such a comparison are presented in Table 3.11. The data reveal that for the 1976-79 period, the last period for which estimates of the capital stock are available, both gross and net enterprise saving fell as a percentage of the capital stock compared to the levels realized in the 1971-75 period. The 1976-79 rates were comparable to the rates that characterized the earlier 1966-70 period in which Yugoslav policy makers first began to worry seriously about the adequacy of enterprise saving efforts. Undoubtedly, the temporary increase in the saving rates relative to capital stock over the 1971-75 period was partly the result of compulsory income controls, the effectiveness of which appeared to drop in the 1976-79 period. lable 3.11: INDICATORS OF SOCIAL SECTOR ENTERPRISE SAVING BEHAVIOR I/ (Economic activities only, in percent) 1966-70 1971-75 1976-79 Gross saving as a share of the replacement capital stock 12.1 14.2 11.8 Depreciation as a share of the replacement capital stock 5.1 6.1 4.9 Net saving as a share of the replacement capital stock 7.0 8.1 6.9 Net saving minus collective consumption funds as a share of the replacement capital stock n.a n.a 4.1 1/ All calculations are based on current prices. Gross saving is defined as depreciation plus allocation to funds, including collective consumption. Source: Statistical Yearbook of Yugoslavia; Laura D'Andrea Tyson, The Yugoslav Economic System and its Performance in the 1970s, Berkeley: Institute of International Studies, 1980.  - 54 - 3.28 Taken in absolute terms, the net enterprise saving rate relative to the capital stock observed during the 1976-79 period is low in the light of reasonable estimates of the net marginal product of capital in the Yugoslav economy. The insufficiency of the enterprise saving effort relative to the productive capital stock becomes even clearer once one takes into account the fact that only about 58X of enterprise net saving during the 1976-79 period was earmarked for additions to the productive capital stock, with the remainder earmarked for various collective consumption expenditures and nonproductive investments. Net saving for productive capital formation was only about 4.1% of the total replacement value of the productive capital stock during this period. Presumably, if Yugoslav firms had been required to set aside earnings in relation to their capital use, the resulting saving rates relative to the productive capital stock would have been noticeably higher than the actual rates observed in recent years.  - 55 - IV. THE 1976-80 PLAN: THE POLICY FRAMEWORK A. Investment Allocation 4.01 As was mentioned earlier, a major instrument to be used in realizing the plan's objectives was the priority system for the allocation of investment. This system, described in detail in Schrenk et. al. (Chapter VIII), rested on the coordination of investment plans by enterprises, sociopolitical communities and banks, through a series of self-management negotiations, self-management agreements, and social compacts. In theory, at least, such coordination was to guarantee the financing of agreed upon projects in priority development sectors, with the funding for nonpriority projects and sectors to be determined residually. 4.02 Compared to the system of investment allocation which it replaced, the new system had two distinct features: first, the identification of priority sectors whose share in the total investment effort was to be guaranteed; and second, the greater use of self-management agreements and self-financing schemes (pooling) among enterprises in lieu of both state and bank intervention as a mechanism to realize the desired distribution of the investment effort. 4.03 Both of these new features were designed to respond to perceived weaknesses in the preexisting investment allocation system. The identification of priority sectors was motivated by concern over structural disproportions in the economy, the most basic of which were deemed to be the relative underdevelopment of the energy and raw material sectors. As noted earlier, a view underlying both the 1971-75 and 1976-80 plans was that the development of these sectors had been retarded by a combination of price controls and commercial policies that favored the processing and manufacturing sectors, and the introduction of a priority allocation scheme reflected the premise that, in the absence of such a scheme, decentralized investment decisions by enterprises and banks would not adequately address domestic capacity constraints in energy and raw materials. 4.04 A full evaluation of the first of these arguments would entail a comparison of the domestic and world prices of tradeable raw materials and tinal goods that goes beyond the scope of this study. Nonetheless, available evidence on price changes over the 1970-79 period can provide some indication of whether the presumed distortions in relative prices increased or decreased during this period. Since the Yugoslav system of price controls has tended to link the domestic price of intermediates, especially energy items, to world price trends, and because world prices of raw materials rose relative to world prices of manufactured goods in the first half of this period, one might expect that these relative price distortions tended to decrease, and the data in Table 4.1 support this expectation.  - 56 - lable 4.1: PRICE TRENDS IN YUGOSLAV INDUSTRY, 1970-79 (ratios of producer prices) 1975/1970 1979/1975 1979/1970 lotal Industrial prices 2.25 1.43 3.29 Investment gooas 1.83 1.38 2.52 Raw materials-intermediates 2.53 1.43 3.62 (of whicb liquid fuels and lubricants) (4.88) (1.77) (8.65) Consumer goods 2.11 1.42 3.00 Agricultural producer prices 2.50 1.80 4.50 Source: Statistical Yearbook of Yugoslavia, various years; also see Statistical Appendix Table A 10.2. 4.05 Over the entire 1970-79 period, the producer prices of intermediate goods rose more rapidly than overall industrial prices, the prices of investment goods, and the prices of consumer goods. The tendency ior the prices of raw materials and intermediates to rise more rapidly than the prices of manufactured goods was characteristic of botb the 1970-75 and 1975-79 periods, although the relative price gain of intermediates was clearly greater during the former period (which coincided with the world boom in raw material prices). All of the priority sectors (with the exception of nonmetallic minerals and chemicals) benefitted from large relative price increases over the entire 1970-79 period, with particularly large increases registered for electricity, coal, and oil. Looking at the period as a whole, it is reasonable to conclude that the presumed distortion in domestic relative prices against raw materials and intermediates was somewhat reduced. 4.06 Despite this shift in the structure of producer prices, it is still possible that the profitability of investment was skewed against the priority sectors, thereby necessitating some form of nonmarket allocation, such as the priority allocation scheme, to ensure that adequate investment did flow to these sectors. Using the CGE model it is possible to determine whether the market indicators of sectoral profitability, resulting from the complex interplay of market forces and policy variables, would have exercised a resource pull in the direction of priority sectors or away from them during the 1976-80 period. 1/ Table 4.2 presents a comparison of 1/ For a full description of the CGE Model, and the definition of sectors in the Model, please see Appendix 1, Volume III. Market profitability rates are measured as the gross marginal value product of capital in each sector. These estimates are generated from estimated production functions and do not correspond to any accounting measure of profitability or net income.  - 57 - Table 4.2: INDICATORS OF RELATIVE MARKET PROFITABILITY 1/ ACROSS SECTORS, 1976-80 (Ratio of sectoral profitability rates to average profitability rate in industry) 1976 1977 1978 1979 1980 1976-80 avg. Industrial average 1.000 1.000 1.000 1.000 1.000 1.000 Electricity .120 .080 .093 .052 .017 .072 Coal 1.203 1.200 1.293 1.421 1.542 1.332 Oil and gas 1.504 1.420 1.267 1.105 1.001 1.259 Ferrous Metallurgy .795 .656 .613 .623 .582 .654 Nonferrous Metallurgy .668 .491 .435 .431 .375 .480 Nonmetallic minerals and construction materials 1.878 2.272 2.344 2.774 2.882 2.430 Mlachinery 1.632 1.740 1.799 1.732 1.525 1.686 Transport equipment and shipbuilding 1.510 1.593 1.702 1.421 1.359 1.517 Electrical equipment 1.978 2.103 2.123 2.066 1.942 2.042 Chemicals and paper 1.170 .895 .804 .769 .772 .882 Textile, leather, rubber, wood and miscellaneous manufactures 1.412 1.509 1.585 1.691 1.963 1.632 Food processing 1.560 1.576 1.396 1.328 1.389 1.450 Services 1.354 1.456 1.302 1.264 1.218 1.319 Construction 4.324 5.648 4.866 5.009 3.979 4.765 Infrastructure .617 .682 .703 .690 .761 .691 Trade 1.960 2.292 2.277 2.448 2.666 2.329 Other productive services 2/ 1.358 1.568 1.690 1.790 2.046 1.690 1/ Profitability in each sector is measured as the gross (i.e. including various fiscal and parafiscal charges) return on the capital stock and it is essentially a measure of the gross marginal revenue product of capital by sector. 2/ Includes crafts, tourism and catering, and other productive personal and professional services. Source: CGE Model for Yugoslavia, Appendix I.  - 58 - profitability rates across sectors, derived from simulation runs of the model. Overall, the results seem to support the view that, in the absence of a priority allocation mechanism, market guided investment decisions would not have pulled investment resources into several of the priority sectors, including electrical energy, ferrous metals, nonferrous metals, chemicals and infrastructure. As the figures in Table 1.3 indicated, these sectors received at least 36% of the total priority investment effort over the 1976-80 period. 4.07 In the priority energy sectors of coal and oil and gas, although market profitability rates remained above the average for all industry, they tended to be lower than rates realized in the manufactured goods sectors, such as machinery, electrical equipment, and consumer goods, including textiles, leather and wood products. The same observation applies to the priority food-processing industry. Only in the priority nonmetallic minerals and construction materials sector was market profitability consistently competitive with rates realized in the manufactured goods sectors. The strong performance of this sector is undoubtedly linked to the effects of the 1976-80 investment drive on the demand for construction materials. 4.08 Finally, it is interesting to compare market profitability rates between services and industry, especially since the priority sector scheme called for an increase in industry's share of total investment resources (from 47.3% in 1971-75 to 51.2% in 1976-80), a share that was already quite high when compared to performance in other economies at similar development levels. 1/ According to the figures in Table 4.2, the overall profitability of services -- defined to include all nonagricultural activities outside of the industrial manufacturing sectors -- was greater than average profitability in industry. Within services, profitability was greatest in construction, trade and other productive services (crafts, tourism and catering, and other personal and professional services). Indeed, these three activities consistently realized the highest profitability rates in the entire economy, yet none of them, with the exception of tourism within the productive service sector, was designated a priority sector. 4.09 Taken together, the results discussed here seem to support the implicit assumption underlying the priority allocation mechanism -- namely, that if investment decisions were guided by decentralized market indicators then the desired reorientation of investment toward energy and raw materials would not occur. While it might be argued that this indicates that the priority sectors chosen were the wrong ones, no such conclusion is warranted, since market indicators of profitability were themselves seriously influenced by underlying distortions in the markets for capital, labor and foreign exchange. Given the sectoral priorities mutually agreed 1/ See IBRD, Raising Productivity in Yugoslav Industry: Some Issues, Report No. 3383-YU, July 1981 p. 13. (Restricted document for internal use only.)  - 59 - upon, even with the shifts in relative producer prices in favor of raw materials and intermediates documented above, some nonmarket mechanism for allocating investment was required. The tension existing between market profitability indicators and sectoral priorities explains in part why the share of the priority raw material and energy sectors in total industrial investment fell short of target. To the extent that enterprises and banks were able to circumvent the new priority allocation mechanism to direct their investment efforts to profitable activities, one would have expected manufactured goods industries to exceed this target, as they did in practice. Finally, tension between market indicators and sectoral priorities, coupled with autonomous pressure for investments in the nonproductive sectors played a role in the excessive growth of investment demand that spilled over into macroeconomic imbalance during the 1976-80 period. For the reasons discussed below, the mechanisms for directing investment resources toward priority sectors were not as effective as planned. As a consequence, the realization of investment priorities resulted in the financing of a larger than sustainable total investment effort through excessive credit creation and foreign borrowing. 4.10 Quite apart from the tension between the price structure and the desired allocation of investment, there remains the question of wbether the priority sectors themselves were correctly chosen. As discussed further below, assigning priority status to entire sectors entailed certain risks, since to do so implied endorsement of a broad spectrum of projects, not all of which necessarily merited special treatment. Nevertheless given the tact that the sector was the unit selected by the Yugoslavs in their system of priority allocation it is relevant to ask whether the selection of sectors was sensible in the light of Yugoslavia's resource endowments. 4.11 It is clear that a full assessment of these choices would need to take account not only of static efficiency, but also such wider-ranging considerations as dynamic comparative advantage, national defense, regional development and externalities of various kinds. It is nevertheless of interest to examine the issue from the limited perspective of efficient capital allocation at shadow, rather than market prices. As explained more fully in Appendix I, Volume III, the CGE model can be used both to calculate measures of relevant shadow prices and to calculate measures of profitability using the shadow pricing assumptions generated by it. Distortions arising from the trade and payments regime have been corrected for by assuming that a flexible exchange rate system operates to equate demand and supply for foreign exchange and by further eliminating all import duties and export subsidies. In order to approximate the existence of a mobile pool of labor in the manufacturing and service sectors, it is assumed that all manufacturing sectors face the same shadow price for labor, which is set equal to the estimated average industrial shadow wage. In the same way, it is assumed that all service sectors face a uniform wage rate, set equal to the estimated shadow wage across all service sectors. This treatment of shadow wages in the manufacturing and service sectors reflects the assumption that the labor market in these sectors is segmented  - 60 - in the medium run. The estimated average shadow wage rate for labor in the service sector is 42% higher than the shadow wage rate for industrial labor. This in part reflects the fact that the highly paid labor in the so-called 'nonproductive services' branches of the economy is included in the estimate of a sbadow wage rate for the service sector. 4.12 The pattern of relative economic profitability that emerges from these calculations is shown in Table 4.3, for 1980. The fact that profitability rates continue to differ by sector is an indication of the fact that no equilibration has been imposed on the allocation of capital. The existence of these profitability differences provides a possible guide to the directions in which capital might be allocated in order to maximize the return to it, the presumption being that, at the margin, capital should be directed to the sectors of higher profitability and away from the sectors of low profitability. Table 4.3: A COMPARISON OF MARKET AND ECONOMIC PROFITABILITY RATES ACROSS SECTORS IN 1980 (Ratios of Sectoral Profitability Rates to Average Profitability Rates in Industry) Market Economic Profitability Profitability Rates Rates Manufacturing and Mining Average 1.000 1.000 Electricity 0.017 0.011 Coal 1.542 1.809 Oil and Gas 1.001 3.161 Ferrous Metals 0.582 0.898 Nonferrous Metals 0.375 0.213 Nonmetallic Minerals and Construction Materials 2.882 3.100 Macn inery 1.525 1.357 Transport Equipment and Shipbuilding 1.359 1.053 Electrical Equipment 1.942 1.819 Chemicals and Paper 0.772 0.642 Textiles, Leather, Rubber, Wood and Miscellaneous Manufacturing 1.963 1.660 Food Processing 1.389 1.777 Services 1.319 1.270 Construction 4.765 3.236 Infrastructure 0.691 0.683 Trade 2.329 4.052 Other Productive Services 1.690 1.870 Source: CGE model estimates.  - 61 - 4.13 The economic profitability indicators indicate a low return to investment resources in the sectors of electrical energy, ferrous metals, nonferrous metals, chemicals, and in infrastructure. The economic return on capital in these branches appears to be substantially below the estimated average economic return to industrial capital. In the case of electrical energy and infrastructure, the existence of substantial externalities makes this measure of economic profitability a poor guide to investment choice and it would be reasonable to support the priority allocation of investment resources to these sectors despite the apparently low rate of return on such investment. A similar argument, however, cannot be easily made in the case of ferrous metals, nonferrous metals and chemicals sectors. Even after correcting for the distorting effects of the trade and exchange regime, the rate of return on capital in these sectors remains substantially below the average industrial rate of return. These results suggest the substantial cost entailed in designating these sectors as priority recipients of investment resources. 4.14 The economic profitability rates, like the market profitability rates, indicate that the rates of return on investment resources were bigher than the industrial average in the priority sectors of coal, oil and gas, nonmetallic minerals and construction materials, and food processing. The results also indicate that the economic profitability rates in these sectors were higher relative to the industrial average than their market profitability rates. By contrast the economic profitability rates of machinery, transport equipment and shipbuilding, electrical equipment and consumer goods (textiles, leather, etc.) are lower relative to the industrial average than their market profitability rates. A comparison of the shifts in rankings between the economic and market profitability measures suggests that trade and labor market distortions acted to depress the relative rates of return in these priority sectors and to increase the relative rate of return on the nonpriority producers of capital and consumer goods. 4.15 Finally, the results on economic profitability suggest that the choice of coal, oil and gas, nonmetallic minerals and construction materials and food processing as priority sectors was economically justified and appropriate. In addition, the high economic profitability in the other productive service sectors, including catering and tourism, supports the priority allocation of investment resources to these activities. 4.16 In addition to the identification of priority sectors, the 1976-80 investment mechanism called for an increase in the role of self-management negotiations and agreements among enterprises in the financing and implementation of investment projects. Self-management agreements between enterprises were to serve as an intermediate coordination mechanism for investment activity, somewhere on the continuum between plan and market.  - 62 - 4.17 Economic theory provides a rationale for the role of interenter- prise negotiations in the area of investment decision-making. 1/ In the absence of information about future prices, the market mecbanism alone does not guarantee optimal decision-making, because the individual enterprise does not have access to the data required to make the most profitable decisions. Without such information, individual investment decisions are not necessarily coordinated and can produce unwanted aggregate effects, such as the duplication of facilities and excess capacity on the one hand, or bottlenecks in certain sectors on the other. Such effects can be avoided if enterprises exchange information about their investment plans prior to undertaking them. Indeed, the exchange of such information is one of the theoretical justifications for indicative planning, which is intended to improve investment decisions by reducing uncertainties about the future behavior of market participants, including the government. 4.18 The negotiation of interenterprise agreements for investment purposes as envisaged by the new law on planning was, however, designed to go beyond this purely indicative role to solve some of the problems of capital mobility and allocation that have characterized the Yugoslav economy through most of the postwar period. 4.19 In the period between the 1965 reform and the reforms of the 1970s, the capital allocation process continued to produce a distribution of capital among regions, sectors and enterprises that was considered suboptimal from several points of view. Interregional optimality was hindered by barriers to capital mobility which encouraged banks, enterprises and local (republican and communal) governments to keep investment resources within local boundaries. Among the well-known and publicized effects of these barriers were the emergence and persistence of "political factories" and the duplication of facilities among republics. The intersectoral allocation of capital was also influenced by artificially low interest rates and price controls. Finally, interenterprise allocations which were affected by both the interregional and intersectoral distribution process were also influenced by the nonprice credit rationing rules of the banking system. 2/ With banks accounting for approximately 48% of total domestic financing for fixed investment over the 1969-72 period, these rationing rules had an important impact on the allocation of funds. Indeed, one of the major reasons for the reforms of the investment allocation mechanism in the first half of the 1970s was concern over "excessive" bank influence over enterprise investment decisions. It was 1/ See Cyrus Ardalan, "Workers' Self-Management and Planning. The Yugoslav Case", World Development, 1980, pp. 623-38 2/ For a fuller discussion of criteria used by financial institutions in appraising investment projects, see Chapter VI.  - 63 - anticipated that both the priority allocation mechanism and the new expanded role for direct enterprise financing and implementation of investment projects would reduce their influence on the allocation of capital. 4.20 Before discussing the actual workings of the investment allocation mechanism over the 1976-80 period, it is important to note that there was little in the new system to guarantee optimality of capital use at the economy-wide level. Optimality at this level would have required that all investment projects be evaluated using some economy-wide indicator of capital scarcity, appropriately adjusted to reflect externalities and resulting divergences between social and private profitability. Ideally, such an indicator would need to guide both the choice of priority sectors and the choice of individual projects in both priority and nonpriority sectors. There was nothing in the new Yugoslav system to guarantee that this condition was met. Even if priority sectors in the plan had been chosen in this way, the mechanism of decentralized enterprise negotiations and agreements in no way guaranteed that individual projects would be appropriately selected. In the absence of a price mechanism for allocating investment funds among competing projects and in the absence of some capital scarcity indicators to evaluate projects, and of an institutional arrangement to make the results of such analysis binding, there was nothing in the new system to guarantee that the relative strengthening of the role of enterprises and the relative weakening of the role of banks would improve the quality of investment decisions made. 4.21 The available evidence from the 1976-80 period suggests that capital misallocation continued to be a significant weakness in the Yugoslav system under the new institutional arrangements. With actual real interest rates falling over time through the combination of constant (or even falling) nominal interest rates and accelerating inflation, nonprice, administrative rationing of capital among alternative projects remained the predominant allocation mechanism. There is no evidence that either project evaluation rules or other criteria used to guide this rationing process in its operation via enterprise agreements or bank or state financing channels acted as a reasonable substitute to evaluate projects according to some standard, economy-wide measure of capital's real scarcity. Instead, anecdotal evidence from banks and planning officials suggests that the criteria used varied from project to project, location to location and time to time. The most one can conclude is that projects in priority sectors tendea to get priority in financing relative to those in nonpriority sectors, even when the projects in the priority sectors were poorly conceived, planned and implemented. In other words, the rationing scheme tended to work to the advantage of projects that were demonstrably priority sector projects but which may have had little else to recommend them.l/ 1/ The Jadral alumina factory is a widely criticized example of a project that received priority sector access to investment financing despite what in retrospect appears to be its inability to survive in contemporary market conditions. See Ekonomska Politika "Bankruptcy as a Way of Rehabilitation?"  - 64 - 4.22 Another general characteristic of the capital allocation process during the 1976-80 period was its continuing fragmentation along regional or even narrower geographic lines. The interest rate structure tended to weaken the interregional mobility of capital since access to financing at a negative real interest rate contained an implicit grant element. Conse- quently, banks, government organizations and enterprises with investment funds preferred to use these for themselves or for their local clients and were reluctant to make interregional loans for even relatively short periods. More generally, the increasing divergence between real interest rates and some appropriate measure of capital scarcity in the system enhanced the role of political and/or administrative factors, including regional or local considerations, in capital allocation decisions. Ironically, as some Yugoslav observers themselves have pointed out, 1/ this situation increased rather than reduced the degree of politicization or statism in investment decisions and aggravated rather than ameliorated underlying regional conflicts about the fair distribution of investment resources. The resulting strengthening of ad hoc, administrative and uncoordinated intervention in investment decision-making not only weakened the efficiency of capital use but also contributed to the increasing administrative costs of running the Yugoslav system. 2/ 4.23 In part, the growing politicization of investment decisions during the 1976-80 period was the consequence of the priority allocation mechanism itself, since, as already noted, realization of priority status was frequently all that was needed to obtain access to investment finance. In part, however, the politicization of investment was also the result of an apparent erosion in the authority of banks vis A vis enterprise borrowers. Discussions with Yugoslav bankers suggest that the ability of the banks to make independent project lending decisions and to monitor the effectiveness of investment projects was indeed circumscribed, as had been intended by the reforms of the investment allocation mechanism. This appears to have been the case throughout Yugoslavia although to differing extents across regions. 4.24 The combination of these influences led to examples of unnecessary duplication of investment facilities across regions, and consequent losses of economies of scale. In the agroindustrial sector for example, eight sugar refineries were constructed, located in each of the republics and provinces (with the exception of Montenegro). The location of these was determined by considerations of regional self-sufficiency, rather than the need to supply the larger geographic market. The overall result has been excess capacity in sugar refining, particularly given the shortfalls in sugarbeet production. A similar pattern was also at work in the 1/ See, for example, Kovacevic et. al., Basic Determinants. 2/ The growing costs of increasing local statism in the Yugoslav economy is also discussed in Kovacevic et al., and was recently criticized in a high level report by Dobroslav Culafic, secretary of the LCY Central Committee Presidium, summarized in the Yugoslav press.  - 65 - construction of slaughterhouses, even where transporting livestock to adjacent republics would have been a lower cost alternative. In the case of electric power generation, while republics and provinces did collaborate in several jointly-owned power station projects at large coal mines and hydropower sites, there was still a pronounced tendency for a region to give priority to develop its own sources of supply, even if cheaper sources were available in another republic's territory. Partly because of this preference, the average size of units installed was much smaller than would be justified by the size of the total interconnected system. Such examples are not restricted to the priority sectors, nor are they wholly attributable to the investment allocation process, since an important enabling factor has been the structure and nature of import protection. The investment allocation system was however important in validating regional aspirations for self-sufficiency, with the associated costs in duplication of facilities and loss of competitiveness that these imply. 4.25 In addition to these microeconomic inefficiencies, the capital allocation mechanism played an important role in growing macroeconomic instability during the 1976-80 period. Low nominal interest rates artificially stimulated investment demand and artificially depressed saving. In theory, quantitative controls on investment finance and its allocation were intended to ration investment demand within the limits of available saving, thus maintaining macroeconomic balance. In practice, and in keeping with past cyclical experience in Yugoslavia, these quantitative controls did not work smoothly or continuously. Instead, the system tended to generate its usual stop-go pattern in whicb rapid increases in investment expenditures in excess of plan targets initially led to excessive increases in domestic credit creation and foreign borrowing. 1/ When the inflationary and balance of payments implications of these increases became unsustainable in the 1979-80 period, investment expenditures slowed dramatically in response to the strengthening of quantitative controls. 4.26 The stop-go nature of the quantitative control system, as in the past, was costly in terms of economic efficiency. During the go phase, projects were financed that would have been rejected under a more consistent rationing scheme. In contrast, during the prolonged stop phase, (which is still in effect as of this writing), potentially profitable projects have been postponed or shelved altogether for want of funds. Given the persistence of high inflation rates, delays in investment completion have considerably increased the cost of projects beyond original financing estimates. In addition, as the following section on the foreign trade and exchange mechanism suggests, the stop-go cycle of the post 1974 period distorted the domestic resource allocation process by making the perceived profitability of import use different for different users at different times and by strengthening the bias toward import substitution. 1/ For a full discussion of the investment-driven nature of macro instability during the 1976-80 period see Kovacevic et. al.  - 66 - Finally, at a macro level, given significant downward rigidities in nominal incomes paid to workers and in prices and given the persistence of infla- tionary expectations, the stop-go cycle aggravated the inflationary bias in the system 1/ and thus increased the severity of the stop or slowdown phase required to reduce inflation and restore balance of payments equilibrium. B. The Foreign Trade and Payments Regime Institutional Developments 4.27 While Yugoslavia retained its official commitment to a liberal trade and payments system over the plan period, the institutional arrangements for formulating and executing policy were substantially altered over the course of the plan. These institutional changes were intended to implement the principles of the 1974 constitution in the trade and payments area. The new framework was articulated in three basic laws, passed by the Federal parliament on March 1, 1977. These were the laws on trade in goods and services with foreign countries, on foreign exchange operations and foreign credit relations, and on conduct of business activities abroad. Of these, the most important for the present discussion was the law on foreign exchange operations and foreign credit relations. The major provisions of these laws came gradually into effect in 1978, but organizational procedures and practices continued to evolve in the light of experience and the the needs of the moment. 4.28 While a detailed review of the new legislation is beyond the scope of this report, its salient features may be sketched, as these are indispensable for an understanding of developments in this period. The major institutional development has been the creation of communities of interest for foreign economic relations (CIFER; known in Yugoslavia as SIZ). These are constituted at the republican and provincial levels, with the regional CIFER providing delegates to a federal level CIFER. The regional CIFER are comprised of delegates drawn from enterprises (organizations of associated labor), organized by industrial branch. They constitute the forum for the articulation and coordination of views and plans on a variety of topics concerned with foreign economic relations. The CIFER are, inter alia, responsible for preparing regional balance of payments projections, for ensuring the consistency of these projections with the foreign trade plans of their constituent organizations, for the administration of export subsidy and foreign exchange retention schemes, and for the allocation of external borrowing rights, consistent with totals set for the region. Of these various duties perhaps the most significant and controversial in this period was the role that the CIFER came to play in distributing available foreign exchange among members; this is discussed more extensively below. 1/ For a discussion of the factors contributing to wage and price rigidities and an inflationary bias in the Yugoslav economy, see Laura D'Andrea Tyson, "The Yugoslav Inflation: Some Competing Hypotheses", Journal of Comparative Economics, June 1977.  - 67 - 4.29 Through their membership in the federal CIFER, the regional CIFER participate in the formulation of the Annual Order on Joint Foreign Exchange Policy which accompanies the annual plan every year, and whicb establishes balance of payments positions for each republic and autonomous province, compatible with the plan for the nation as a wbole. In instituting the new system of CIFER it was thougbt that by transferring responsibilities for foreign excbange balance from state bodies at the iederal level to regional self-managing bodies, greater discipline would be imposed on importing enterprises through the closer and more direct link with the realized export performance of the republic or autonomous province; also that when emergency measures to limit imports needed to be undertaken this could be done less arbitrarily and disruptively under the auspices of an enterprise-based body such as the CIFER, rather than being imposed by the state. 4.30 A further feature of the new legislation was its treatment of foreign exchange retention rights. Consistent with the underlying premises of the system, ana in recognition of the fact that intermediate producers contribute to exports without being able to realize foreign exchange directly, Article 67 of the new legislation allowed OALs to conclude self-management agreements for the distribution of foreign exchange earnings under the auspices of the regional CIFER. Since there were practical difficulties in making this mechanism the sole source of foreign exchange for importers, provision was also made for the regional CIFER to allocate foreign exchange to eligible importers in accordance with current balance of payments policy. Incentive Framework 4.31 Given the relatively poor export performance of the 1976-80 period described in Chapter II and the ensuing balance of payments difficulties of the later years of the plan, an important question about the changing foreign trade and payments regime is the direction and strengtb of the biases that it created for producers to produce for the home market as versus export markets. In this context, the bias of the incentive system refers to the effects that the system has in altering the relative profitability of producing for the home market rather than for exports as compared to a free trade situation. In the Yugoslav system, as elsewhere, the overall trade bias reflects the interaction between numerous policy instruments, including the official exchange rate, the structure of tariffs, explicit and informal quotas on imports, export subsidies and selective credit policies. While it has not proved possible to take all of these influences fully into account, the CGE model does provide a framework within which the allocative effects of these instruments can be analysed. 4.32 The structure of Yugoslavia's commercial policies was analyzed at length in an earlier report 1/, and is described in some detail below in 1/ Yugoslavia: Export Performance and Policies.  - 68 - Chapter VI. The earlier analysis concluded that while no very firm judgment could be made on the trade bias of the official commercial policy regime, it was likely that tariff and quota protection as provided for in Yugoslavia's trade legislation slightly exceeded explicit subsidies provided to exporters, creating only a modest bias against exports. The report noted, however, that this finding contradicted widespread assertions of Yugoslav enterprises that it was more profitable to sell at home than to export over much of the 1976-80 period. The report concluded that a substantial antiexport bias existed, generated through the operation of informal rather than formal mechanisms of commercial policy; in particular through increased resort to quantitative restrictions (QRs) as an instrument of balance of payments management. The finding of a strong bias against exports is consistent with the highly cyclical and erratic pattern of exports over the period, a pattern indicating that Yugoslav firms turned to foreign markets only when the domestic economy was relatively depressed, or when direct administrative pressure was exerted upon them to do so, as in 1976 and again in 1980-81. It also conforms with the view of many Yugoslav economists that the system had a massive bias against exports over much of this period. 1/ 4.33 The conclusion that QRs were of substantial importance over this period is corroborated from a variety of sources. Table 4.4 provides information on the formal quota regime, and it is apparent from this that there was a significant decrease in fully liberalized imports in 1976, from 54% of total imports in 1975 to 44% of total imports in 1976, and that this decrease persisted for the remainder of the period. Increased resort to QRs in 1976 is also supported by the ratios of imports to domestic supply in current prices for the years of the plan shown in Table 4.5. The picture that emerges is that the reduction in imports initiated in the growth pause of 1975-76 was by and large maintained in the 1976-79 period. 4.34 Although it is possible to argue that this decline was the consequence of effective import substitution and the changing relative prices of imports and domestic substitutes, this conclusion seems unlikely given the suddenness and consistency of the decline throughout the economy, the growing reliance on quantitative restrictions suggested by the data in Table 4.4, and the fact that the bulk of the import substitution projects begun in 1976 were not completed, if at all, until several years later. It seems reasonable to conclude that the apparent decline in import dependence that marked the beginning of the 1976-80 period and that appears to have persisted thereafter was in large measure the result of the greater use of quantitative restrictions. 1/ For a discussion of the disincentives against exports in the Yugoslav economy over the 1976-80 period, see for example Mate Babic, Strukturna Analiza, op. cit.; Jose Mencinger, "Zatvaranje Jugoslavenske Privrede u Sistemu Zastite", Ekonomski Institut, Pravne Fakultete, Ljubljana, 1979; Ljubomir Madzar and Zoran Popov, Stabilizacija i Razvoj Jugoslavije do 1985, Institut Ekonomskih Nauka, Belgrade, 1981 (Mate Babic, Structural Analysis, op. cit.; Jose Mencinger, "Closing in of Yugoslav Economy in the Protection System", Economic Institute of the School of Economics, University of Ljubljana, 1979; Ljubomir Madzar and Zoran Popov, Stabilization and Development of Yugoslavia through 1985, Institute of Economic Sciences, Belgrade, 1981).  - 69 - Table 4.4; IMPORTS BY IMPORT REGIME, 1971-80 (Value share of total imports, in percent) 1971 1973 1975 1976 1977 1978 1979 1980 A. Fully liberalized (LB) 28.7 51.6 54.3 43.7 45.0 46.3 42.5 42.4 B. General foreign exchange quota (GDK) 45.7 21.2 17.2 17.4 18.8 0.0 0.0 0.0 C. Value quotas (DK) 20.3 20.6 4.0 4.3 4.4 18.2 18.5 16.2 D. Quantity quotas (RK) - 17.5 17.5 16.9 14.7 21.8 23.3 20.2 E. Licensed (D) 5.3 6.5 7.0 17.7 17.1 13.7 15.7 21.3 Percent subject to quota or licensing (B+C+LI+E) 71.3 48.4 45.7 56.3 55.0 53.7 57.5 57.6 Source: Statistical Appendix Table A 3.10. 4.35 The resulting slowdown in imports occasioned by these import restrictions is apparent in various estimates of average import elasticities over the 1976-80 period compared to the 1971-75 period. On an aggregate level, as the figures in Table 4.6 reveal, the elasticity of real merchandise imports with respect to real gross material product dropped by more than 50%, from a 1971-75 average of 1.08 to a 1976-80 average of 0.49, below the 1976-80 plan target of 0.55. This dramatic reduction in the aggregate figures is reflected in figures for intermediate imports and capital equipment imports which together accounted for about 88 percent of all imports during the period. 4.36 As the data in Table 4.6 indicate, the elasticity of raw material and intermediate imports with respect to real gross material product droppea from a 1971-75 average of 0.97 to a 1976-80 average of 0.79. The drop is somewhat smaller -- from a 1971-75 average of 0.72 to a 1976-80 average of 0.65 -- if the elasticity is measured with respect to real gross material product in industry and smaller still -- from a 1971-75 average of 0.70 to a 1976-80 average of 0.67 -- if the elasticity is measured with respect to the physical volume of industrial production. The differences among these figures are attributable to several factors, including: (i) a reduction in the contribution of the growth of industrial material product to the growth of overall material product in the 1976-80 period, offset by an increase in the contribution of the growth of construction and public utilities, botb of which are less intensive users ot raw material imports than manufacturing; (ii) a quantity rationing scheme that favored industrial users of raw material and intermediate imports over nonindustrial users; and (iii) a more efficient use of material inputs in industry, as a result of which the elasticity of intermediate imports with respect to real industrial value-added fell more sharply than the elasticity of intermediate inputs with respect to the gross volume of industrial production.  - 70 - Table 4.5: RATIO OF IMPORTS TO SUPPLY AVAILABLE FOR DOMESTIC USE, 1972-1979 1/ (Current price estimates, in percent) 1972-75 1975-76 1976-79 (average) (average) (average) Agriculture 6.7 7.3 7.3 Forestry 13.3 9.8 9.9 Coal 19.9 18.8 17.8 Oil 58.6 62.6 60.2 Ferrous metals 32.7 22.6 20.8 Nonferrous metals 22.3 14.2 13.3 Nonmetallic minerals 40.6 32.6 30.2 Construction materials 1.3 1.8 1.8 Metal products 41.2 36.6 35.7 Shipbuilaing 31.9 57.7 41.2 Electrical equipment 26.3 24.9 24.7 Chemicals 42.7 36.3 36.7 Wood products 5.6 3.5 3.3 Paper 18.8 14.2 12.8 Textiles 15.9 7.7 7.4 Leather 9.3 4.0 4.4 Rubber 35.0 28.5 25.3 Food processing 6.4 6.4 5.1 I/ The ratio of imports to domestic supply available for domestic use is measured as the dinar value of imports divided by the gross dinar value of production less the dinar value of exports. All values are calculated in current prices. The dinar values of imports and exports are calculated by applying the period average exchange rate to the dollar values of imports and exports in each year. Hence, the dinar values are exclusive oi tariffs and other import charges but inclusive of any direct export subsidies that affect the dollar values of exports. All calculations are based on gross value of production figures and dollar import and export figures contained in the Statistical Yearbook, various issues.  - 71 - Table 4.6: INDICATORS OF IMPORT ELASTICITI, 1971-1980 I. Aggregate Import Elasticities 1/ 1971-75 1976-80 Average annual rate of growth: 2/ Real merchandise imports 6.4 2.8 Real gross material product 5.9 5.7 Real gross material product in industry 7.9 6.9 Average annual import elasticity witb respect to: Real gross material product 1.08 0.49 Real gross material product in industry 0.81 0.39 II. Intermediate Goods Import Elasticities Average annual rate of growth: Real imports of raw materials and semi-manufactures 5.7 4.5 Industrial production index 8.1 6.7 Average annual import elasticity of raw materials and semi-manufactures with respect to: Real gross material product .97 .79 Real gross material product in industry .72 .65 Industrial production .70 .67 III. Capital Goods Import Elasticities Average annual rate of growtb: Real imports of capital goods 1/ 10.7 -3.3 Real gross investment in fixed capital 2/ 7.2 7.0 Average annual import elasticity of capital goods imports with respect to Real gross material product 1.81 -0.58 Real gross investment in fixed capital 1.49 0.47 1/ Calculated as the ratio of the average annual rate of growth of real imports to the average annual rate of growtb of the aggregate indicator with respect to which the elasticity is measured. 2/ Average annual rates of growth of real imports in aggregate and by category are calculated from dollar values of mercbandise imports by end-use and import price indexes by end-use (1975=100). Source: Statistical Appendix Table A 3.6 and information provided by the Yugoslav authorities.  - 72 - 4.37 An even more dramatic decline in import dependence is suggested by the figures on the elasticity of capital equipment imports with respect to real gross material product or real gross investment. Both figures reveal a sharp drop from a positive elasticity (in the range of 1.49 to 1.81) for the 1971-75 period to a negative elasticity of (-.47) to (-.58) for the 1976-80 period. Because of the difficulties involved in measuring changes in import prices of equipment over time -- difficulties due in large part to the highly specialized and variable character of equipment imports -- these figures must be interpreted with caution. Nonetheless, given the magnitude of the changes involved, it is reasonable to conclude that there was a significant drop in the elasticity of equipment imports over the 1976-80 period. This drop is all the more dramatic given the investment boom that occurred in Yugoslavia during the same period. Foreign Exchange Allocation 4.38 In the presence of persistent foreign exchange shortages and particularly after the creation of the regional CIFER in 1978, the Yugoslavs resorted to a complicated set of rationing devices to allocate available foreign exchange among competing users. In essence, two major forms of rationing were in operation in varying degrees over the period. Simple quantitative restrictions on either the allowable quantity of imports or on the allowable foreign exchange made available for a given type of import or to a given class of importer represent what will be called "fixprice" mechanisms for allocating import rights. The distinguishing characteristic of a fixprice rationing mechanism is that, although the actual quantity of imports allowed to a domestic user falls short of the amount he would wish to purchase, the restricted quantity that he is permitted to purchase is provided to him at an official price, determined by the world price, the official exchange rate, and the official tariff and import tax structure. To take a simple example, enterprises importing crude oil are provided the foreign exchange that they need to import a given quantity of oil at the official exchange rate. They pay import duties on this import (in dinars) at the officially established import duty rate. The landed price of oil for them in dinars is thus the world price of oil (including transport costs) translated into dinars at the official exchange rate plus any import duties and other trade related taxes they are called upon to pay. In the same way, all the formal foreign exchange and quota restrictions used by the Yugoslavs to control imports are fixprice mechanisms, because the right to import under each of these restrictions is a right of access to the foreign good at its official dinar equivalent, tariff-inclusive, price. 4.39 While these fixprice mechanisms for allocating foreign exchange remained important throughout the 1976-80 period, there is also anecdotal evidence that following the 1978 changes in the foreign exchange allocation system there was growing use of a "flexible price" or "premium rationing" mechanism in 1979 and 1980. The distinguishing feature of this "flexprice" mechanism was that certain potential users of foreign exchange were required to pay a premium for foreign exchange, above the official exchange rate, in  - 73 - order to obtain the foreign exchange they needed. The post-1978 institutional arrangements allowed earners of foreign exchange to negotiate self-management agreements to circulate their foreign exchange earnings to other deficit users of foreign exchange. This introduced the possibility that such agreements implicitly or explicitly could value the foreign exchange being transferred at a rate which represented a premium over the official exchange rate. Anecdotal evidence suggests that a premium valuation of foreign exchange above the official exchange rate increasingly became the norm in such contracts in 1979 and 1980, reflecting the growing divergence between the official exchange rate and the underlying market clearing exchange rate. 4.40 In the case of such premium valuation of foreign exchange, an importer forced to pay a premium on the foreign exchange he obtains would evaluate the cost to him, in dinars, of imports at a price which reflected this premium. A situation where premium rationing applied uniformly to all importers would be equivalent to a uniform higher exchange rate, on the import side. It appears that both quantitative (fixprice) and premium (flexprice) rationing schemes coexisted in Yugoslavia in 1979 and thereafter. It seems, however, that the premium market worked only partially to equate demand and supply of foreign exchange, constrained as it was by the need for enterprises to arrive at long-term cooperation agreements in order to legitimize the transfer. The coexistence of both modes of rationing implies that different users of the same imports paid different dinar prices for those imports, depending on whether they had access to foreign exchange at the official or the premium-ridden price or some combination of the two. It seems likely that the resulting distortions reduced the overall efficiency with which foreign exchange was used in the system in comparison with what would have occurred under either a more flexible exchange rate system or a system of market-determined premia paid by all importers. 4.41 Relying on the assumption that fixprice rationing was the pre- dominant method of rationing foreign exchange in the 1976-78 period, after which flexprice methods became more important, the CGE model solves for the average quantity and premium rationing rates presented in Table 4.7. 1/ These results indicate that the greatest degree of quantity rationing occurred in 1976 and 1978 when importers were able to obtain only about 80% of their desired imports. After 1978, the role of quantity rationing declines, although it does not disappear, and the average premium rate expressed as the markup on imports valued at world prices rises from 0% in 1978 to 12.3% in 1979 and 17.9% in 1980. 1/ Please see Appendix I for details.  - 74 - Table 4.7: AVERAGE QUANTITY RATIONING AND FOREIGN EXCHANGE PREMIUM RATES, 1976-80 1976 1977 1978 1979 1980 Average quantity rationing rate 1/ .797 .888 .804 .843 .901 Average premium rate 2/ .000 .000 .000 .123 .179 1/ The quantity rationing rate is defined as the ratio of actual imports to desired imports. The average quantity rationing rate in 1976 is based on exogenously specified sectoral rationing rates and the actual structure of imports in world prices. In the following years, the average quantity rationing rate reflects exogenous assumptions on the severity of rationing over time with respect to imports of raw materials, capital goods, consumer goods and services. 2/ See text for definition. 4.42 The existence of substantial unsatisfied demand for foreign exchange at the official exchange rate creates substantial rental income in the system. Two situations may be distinguished. Firstly there is the group of importers who are beneficiaries of the fixprice rationing techniques described above; i.e. they are successful at obtaining foreign exchange at the official rate, though usually in an amount less than they would desire. The fact that they enjoy this access implies a transfer payment to them, a transfer which is realized in the profitability of their production in comparison with a flexible exchange rate alternative. Second, there is the group of importers (some of whom may be the same as those in the first group) who are able to meet their unsatisfied demand for foreign exchange through bidding for foreign exchange, at a premium exchange rate, from those who happen to have foreign exchange in excess of their needs. As noted, it is assumed that because of institutional imperfections in such secondary transactions in foreign exchange, the premium rate does not rise to a level adequate to extinguish all unsatisfied demand, and it is this level of unsatisfied demand which is reflected in the average quantity rationing rate of Table 4.6. 4.43 The existence of this scarcity rent coupled with the frictions in the premium 'market' for foreign exchange are likely to encourage rent- seeking behavior. Such rent-seeking behavior has a real cost associated with it, as resources that might otherwise be used in production are redirected to such activities as lobbying for quantity allocations in communities of interest, negotiating agreements with enterprises for the distribution of foreign exchange and the like. Even if actual rent-seeking behavior does not develop to the full degree that is possible, it is of some interest to obtain measures of the total rental income that arises in the economy.  - 75 - 4.44 The calculations of rent by sector of final production are based on the assumption that, in general, the rents associated with the scarcity of foreign exchange accrue to the first round users of imports. 1/ Given the institutional arrangements for trade in Yugoslavia where enterprises are directly allocated foreign exchange and import licenses to cover their needs, potential rents accruing to an output sector are calculated as the sum of rents associated with its imports of inputs from all sectors of origin (including its own). 4.45 The magnitude of rents calculated in this way and expressed as a percentage of value-added for each sector is given in Table 4.8. The results indicate, as expected, that the overall degree of distortion in the economy, as measured by the scale of import-related rents, increased significantly in 1979 and 1980 as a result of the developing foreign exchange crisis, the increasing overvaluation of the dinar, and the greater severity of import rationing. The results also indicate that rental income as a percentage of value-added is greater for the more import-dependent sectors: that is, for the sectors that depend most on purchases of imports of raw material and capital inputs for production. Thus, rental income is greater overall for industry than for either agriculture or services, and within industry, it is greater for the more import-dependent sectors, such as electricity, oil and gas, ferrous metals, transportation equipment, electrical equipment, and chemicals. 4.46 The results also suggest that the potential for rent-seeking behavior in the system due to the rising value of rents were quite dramatic over the whole period. With all good will and social pressure, the power of these incentives must have encouraged rent-seeking behavior in a variety of forms, consistent with the institutional constraints facing economic actors. For example, rent-seeking undoubtedly made the process of reaching self-management agreements about foreign exchange and import rationing more time-consuming and divisive than would otherwise have been the case. Even if all of this rent was not dissipated in rent-seeking activities, the scale of the gains to be realized through privileged access to foreign exchange was clearly very large, and provides an explanation for the fierceness of the interregional and interenterprise debates on the rules for allocating foreign exchange in this period. Export Incentives 4.47 The discussion so far has concentrated on how the mechanisms for allocating foreign exchange influenced the behavior of importers during the 1976-80 period. These same mechanisms, together with other additional policy interventions, also influenced the behavior of exporters. As in the past, so in the 1976-80 period, the Yugoslavs employed at least three major 1/ See also the discussion of rent incidence in Appendix I.  - 76 - Table 4.8; RENTAL INCOME FROM FOREIGN EXCHANGE RATIONING AS A PERCENTAGE OF SECTORAL VALUE ADDED, 1976-80 1976 1977 1978 1979 1980 Agriculture 2.19 1.04 1.77 2.91 2.98 Electrical energy 20.03 13.96 21.72 29.64 33.71 Coal 8.73 5.37 6.85 8.37 8.27 Oil and gas 39.38 13.57 18.21 55.05 122.32 Ferrous 37.60 18.19 23.83 34.15 40.43 Nonferrous 18.34 11.36 16.62 21.45 23.17 Nonmetallic minerals and construction materials 9.52 4.31 5.72 6.62 7.52 Machinery and fabricated metals 15.84 9.27 12.92 16.07 17.59 Transportation equipment and shipbuilding 27.27 16.11 22.40 29.75 30.69 Electrical equipment 21.03 12.29 17.57 23.24 25.20 Chemicals and paper 25.73 14.87 19.08 29.04 30.41 Textiles, leather, rubber, wood and miscellaneous manufacturing 10.64 5.90 9.29 11.20 10.73 Food processing 8.12 4.84 9.71 16.05 16.64 Construction 7.23 3.11 5.04 6.09 7.69 Infrastructure 9.17 5.30 7.44 9.54 9.21 Trade 1.96 1.00 1.52 1.79 1.80 Crafts and other productive services 3.57 1.98 2.95 3.83 3.76 Primary (agriculture and food processing) 3.29 1.69 3.09 4.85 4.89 Industry 17.86 9.78 13.70 17.90 19.37 Services 1/ 4.57 2.48 3.69 4.46 4.47 Average 8.29 4.33 6.39 8.24 8.59 1/ Includes construction, infrastructure, trade, crafts and other productive services. Source; CGE Model for Yugoslavia, Appendix I.  - 77 - policies to promote exports.l/ First, they continued to grant selective credit at favorable interest rates for the preparation and financing of exports. An estimated 54% of total selective credit was used for this purpose between 1976 and 1980. In addition, projected export performance was supposed to be one of the criteria guiding the rationing of investment resources during the period. Unfortunately, there is no way to assess the significance of this export promotion mechanism, because there is no evidence on the share of investment resources allocated to export promotion projects. 4.48 Second, exports were stimulated by various duty drawback and reimbursement schemes including other favorable fiscal treatment, such as reductions in various taxes for firms exporting a certain percentage of their output. In total, the resources devoted to export promotion from such fiscal schemes amounted to an estimated 56% of total import duties during the 1976-80 period. Finally, these direct fiscal incentives for exports were bolstered by an export retention scheme that gave exporters discretionary control over a varying fraction of their foreign exchange earnings during the period. The incentive effects of this scheme depended in turn on the workings of the fixprice and flexprice mechanisms for allocating foreign exchange discussed earlier. 4.49 In practice, the Yugoslavs combined both flexprice and fixprice rationing witb varying retention rates and varying limitations on the discretionary use of retained foreign exchange. Because of the many ad hoc, administrative interventions that limited the discretion of YugosTav exporters over the use of their retained foreign exchange, it is necessary to distinguish between formal and effective retention rights. The provisions of the new law on foreign exchange operations and foreign credit relations, as described earlier, gave earners of foreign exchange the right to dispose of it freely via self-management agreements negotiated through the CIFER. This in theory implied a substantial increase in retention rates: from an average of 20% in the 1976-78 period up to 100% under the new system. Higher formal retention rates, combined with increases in potential rental income from both premium and quantity rationing in 1979 and 1980, would have implied a higher value being imputed to the retention schemes by exporters and hence a higher incentive to exports arising from such schemes. The enlargement of formal retention rights and the growing implicit subsidy to exports that they represented, however, were not matched by an equal expansion of effective retention rights, given the growing ad hoc limitations on enterprise use of foreign exchange that developed iWnresponse to the foreign exchange crisis. Such limitations tended to introduce uncertainty on the actual retention levels that would be permitted and hence to reduce the perceived implicit incentive for exports associated with such rights. 1/ For a fuller description of export incentive policies and their administration see Chapter VI.  - 78 - 4.50 Increasingly over the course of 1979 and 1980, while the right to import was tightly limited to export earnings, earnings over and above those required for the enterprise's essential imports were subject to reallocation to deficit importers through the intervention of the CIFER. Under such circumstances, the ability of exporters to negotiate the disposition of foreign exchange earnings in excess of their needs with deficit importers was increasingly restricted. As a result, not only were the incentive effects of the retention scheme lower than they would otherwise have been but they also varied considerably across enterprises, sectors and regions of the economy, producing a plethora of effective exchange rates and distorting the allocative efficiency of economic decisions. 1/ Using the CGE model it is possible to obtain an approximate measure of the subsidy to exports resulting both from various fiscal promotion schemes, including the tariff rebate program and the retention scheme. 2/ 4.51 In order to estimate the subsidy rate however, it is necessary to have an estimate of the effective retention rate -- that is, the share of total export earnings that were under the effective (as opposed to formal) discretionary control of exporters. The estimates below assume that the effective retention rate rose from 20% in 1976 to 30% in 1977 and then fell back to 20% in 1980, as a result of increasing administrative allocation of foreign exchange use in response to the growing foreign exchange shortage. The value to the exporter of foreign exchange retained at these rates is measured by the imputed rental income realizable by the exporter in his import purchases. The subsidy rate on exports arising from the retention scheme is estimated as this rental income expressed as a percentage of total export earnings. While this method of estimation in principle allows for sectoral differences in the implicit subsidy rates on exports, in the absence of information on such differences, the CGE model only measures an average, economy-wide subsidy rate. As far as formal subsidy programs are concerned, the CGE model assumes that, througbout the period, formal subsidies in the amount of 55% of customs tariff receipts were made available in the form of tariff rebates and other fiscal incentive programs. 1/ The OECD Economic Survey: Yugoslavia, 1981 contains a discussion-of how the tight links between export performance and import rights tended to produce a variety of different shadow exchange rates. 2/ Excluded from this measure are the subsidy effects conferred tbrough selective credit policies. Although such effects may have been substantial, they are not quantifiable within the CGE model as currently specified.  - 79 - 4.52 On these assumptions, Table 4.9 contains estimates of an average economywide subsidy rate on exports over the 1976-80 period. The jump in subsidy rates in 1978 and 1979 is attributable to the increases in retention rates that are assumed to have occurred in those years, combined with increases in the imputed value of the retention scheme, as potential rental income from quantity and premium rationing grew. The drop in 1980 reflects both a decline in the effective retention rate as a result of growing administrative restrictions on the use of retention rights, and the abolition of the import surcharge that reduced the formal subsidies paid from duty rebates. The 1980 decline in overall subsidy rates is consistent with reports of a temporary reduction in the scope of export subsidy programs following the major devaluation of June 1980. Table 4.9: ESTIMATED AVERAGE EXPORT SUBSIDY RATE, 1976-80 1976 1977 1978 1979 1980 24.4% 19.9% 25.9% 28.5% 17.4% Sourcez CGE Model for Yugoslavia, Appendix I. Trade Bias 4.53 Combining the foregoing analysis of the import rationing and export subsidy schemes, it is possible to examine the direction and extent of the bias in incentives for exporting and import substituting both within and among sectors of production. The measure of bias used in this discussion is the domestic resource cost (DRC) of earning or saving a unit of foreign exchange in each sector, defined as the marginal direct and indirect domestic cost of primary factors of production (measured in domestic currency) divided by the net foreign exchange earnings from a unit of production. The foreign exchange earned is either earned directly by exporting the good or indirectly from using it as a substitute for imports. In either case, foreign exchange earnings are measured net of direct and indirect intermediate import costs. As calculated, the sectoral DRCs have the same units as an exchange rate and can be compared to economy-wide exchange rate indicators. 4.54 Within a sector of production, the DRCs for exporting and for import substituting will be the same only if the effective exchange rate (between domestic and world prices) is the same for both imports (and hence import substitutes) and exports in that sector, and this in turn will be the case only when the nominal exchange rate augmented by tariffs and import taxes on the import side equals the nominal exchange rate augmented  - 80 - by direct and indirect subsidies on the export side. 1/ Thus, divergences in these DRCs provide a measure of the relative incentives generated by the trade regime for production for the home market rather than for exports. Across sectors, differences in measured DRCs can result both from sectoral differences in the effects of the trade regime -- tariffs, quotas, premia, subsidies and retention schemes -- and from sectoral differences in the domestic costs of equivalent inputs. Strictly speaking, the effects of these latter differences should be removed from DRC calculations to get precise measures of the intersectoral effects of the trade regime. This is usually accomplished by evaluating DRCs using shadow price estimates of domestic inputs. Because of the many difficulties involved in obtaining such shadow price estimates and because of their approximate nature, Table 4.10 presents two alt-rnative sets of DRC calculations, one set using the actual prices of inputs and the other set using shadow prices of inputs derived from the CGE model. 2/ Although some of the sectoral rankings differ between the actual and shadow price estimates, the overall conclusions to be drawn from the DRC calculations are the same for both sets of estimates. The most important appear to be the following. 4.55 First, within almost all sectors of production, the DRC at market prices for import substitution was substantially higher than the DRC for exports during the 1976-80 period. This was especially true for manufacturing, which in 1980 (the year on which the calculations are based) accounted for about 65% of dollar earnings from exports of goods and services and about 86% of dollar expenditures on imports of goods and services. Within manufacturing, the divergence between the DRC for import substitution and the DRC for exports was greatest in nonferrous metals, machinery, transportation equipment and shipbuilding, electrical equipment, chemicals, textiles, leather, rubber and other miscellaneous manufacturing sectors. These results attest to the bias against exports that prevailed in the system by the end of the period as a result of the trade and payments regime. 4.56 Second, the wide dispersion of DRC rates across sectors, especially on the import substitution side, points to the sectoral biases introduced in the system by the trade regime. In a neutral trade regime, with uniform tariffs and subsidies and with a flexible exchange rate, the DRCs based on shadow price estimates of input costs would be equal for 1/ The DRC measure is thus formally equivalent to more conventional measures of trade bias, which compute the ratio between effective exchange rates on exports and import substitutes. 2/ The DRC estimates based on actual input prices are properly speaking estimates of effective protection and not estimates of domestic resource costs, since the latter term usually presupposes the shadow pricing of inputs.  - 81 - Table 4.10; ESTIMATED SECTORAL DOMESTIC RESOURCE COSTS, 1980 i/ Market Prices 2/ Shadow Prices 3/ Export Import Ratio Export Import Ratio (1) (2) (2O1) (3) (4) (4 3) Agriculture 29.1 33.6 1.15 24.7 25.3 1.02 Electricity 0.0 0.0 - 0.0 0.0 - Coal 28.3 32.0 1.13 23.6 28.5 1.21 Oil and Gas 14.2 31.0 2.18 7.9 32.2 4.07 Ferrous Metallurgy 22.8 35.0 1.54 17.4 37.4 2.15 Nonferrous Metallurgy 25.6 43.4 1.70 17.1 45.2 2.64 Nonmetallic Minerals and Construction Materials 26.7 33.1 1.24 17.3 25.9 1.50 Machinery 25.3 51.7 2.04 17.1 45.7 2.67 Transportation Equipment and Shipbuilding 23.9 59.5 2.49 14.5 49.4 3.41 Electrical Equipment 25.3 54.7 2.16 17.1 46.7 2.73 Chemicals and Paper 22.7 42.9 1.89 11.8 41.8 3.54 Textiles, Leather, Rubber, Wood and Miscellaneous Manufacturing 22.5 63.3 2.81 19.0 55.3 2.91 Food Processing 26.4 31.9 1.21 18.8 26.3 1.40 Construction 26.1 0.0 - 16.6 0.0 - Infrastructure 27.8 29.7 1.07 26.4 32.2 1.22 Trade 19.3 0.0 - 13.5 0.0 - Crafts and other Productive Services 27.1 35.2 1.30 20.7 31.1 1.50 1/ DRC, measured in Yugoslav dinars, is defined as the total (direct and indirect) cost, in terms of domestic labor and capital, of earning or saving one dollar of foreign exchange through exporting or import substitution. 2/ Equivalent to effective rates of protection. 3/ Shadow prices for labor were estimated using the CGE model for Yugoslavia. See Appendix I for a discussion of the assumptions used in calculating shadow prices.  - 82 - exports and import substitution within each sector and across all sectors. 1/ By contrast, the DRCs based on the actual distorted trade and exchange regime of the Yugoslav economy sbow wide sectoral divergences. Sector-specific quantity rationing of imports and sector-specific premium rates result in particularly large intersectoral differences in the cost of earning a dollar of foreign excbange via import substitution. 2/ In general, the highest DRCs for import substitution are in the sectors in which import restrictions were greatest, such as machinery, transport equipment and shipbuilding, electrical equipment and consumer goods (textiles, leather, rubber, wood and miscellaneous manufacturing). 4.57 Third, since the DRCs for exporting and import substituting in each sector measure the effective exchange rate for that activity in that sector, these DRCs can be compared to estimates of an appropriately defined shadow exchange rate to determine whether it is desirable or economically (as versus financially) "profitable" for a particular sector to engage in exporting or import substituting. In general, an activity with a DRC lower than the shadow exchange rate should expand since it can earn foreign exchange at a domestic resource cost that is lower than the economy-wide opportunity cost measured by the shadow exchange rate. In order to determine the social profitability of exporting or import substituting in a particular sector, it is first necessary to estimate the shadow exchange rate to be used. 4.58 Various simulation runs of the CGE model for Yugoslavia discussed below suggest a 1980 shadow exchange rate ranging between a lower bound of 25 dinars to the dollar and an upper bound of 30 dinars to the dollar. The experiments on which this range is based all assume a flexible exchange rate regime, although they differ in their other assumptions, including assumptions about sectoral tariff and subsidy rates, sectoral indirect taxes, the demand responsiveness of exports and the use of actual or shadow input prices. Because of the difficulties involved in defining and measuring the shadow exchange rate, the discussion and conclusions here are based on a range of estimates rather than on a single estimate. Using this range as a guide, the DRC results in Table 4.10 yield some conclusions about the economic profitability of exporting versus import substituting in different sectors. 1/ Strictly speaking this is true only when there are no sectoral biases in the pricing of factors, so that all sectors face the same input prices, and when there are no direct taxes. 2/ The fact that the dispersion of sectoral DRCs for exports is smaller than the dispersion of sectoral DRCs for imports is in part the result of assuming standard direct export subsidy rates and retention quotas across all sectors. In reality, these rates and quotas differ across sectors so that the actual DRCs for exports probably differ more across sectors than the results in Table 4.10 suggest.  - 83 - 4.59 Perhaps the most striking conclusion is that exporting activities in all sectors are unequivocally economically profitable in the sense that the DRCs for these activities (whetber measured at shadow or market input prices) are less than or equal to the entire range of estimates of the shadow exchange rate. Moreover, in the large majority of sectors, the DRCs from exporting are substantially below tbis, indicating room for an economically profitable expansion of exports. In contrast, import substituting activities appear to be economically profitable in only a few production sectors, including agriculture, coal, nonmetallic minerals and construction materials, and food processing. In all other sectors, the DRCs for import substituting activities lie above the range of estimated sbadow exchange rates, sometimes substantially so. The major reason for this is that the user price of imports (including premia paid on foreign exchange) is mucb bigher than the price given by the official exchange rate thereby greatly shielding production of these goods for the home market. The result is an across-the-board bias against exporting activities and in favor of import substituting activities that shows up clearly in the results in Table 4.10. In industry, the bias leads to an average DRC for import substitution which is about 40% above the upper range estimate of the shadow exchange rate. These results underline the costs to Yugoslavia both of its longer-term import substitution efforts, and of dealing with the emerging foreign exchange crisis during the 1976-80 period through import rationing. The emergence of very large import premia caused major distortions in the incentives facing individual sectors and a substantial overall bias in incentives against exporting. The result was a price structure which inhibited exports and encouraged sales on the domestic market to the extent that the cost to the economy of saving a dollar through import substitution was raised 40% above the cost that would have been necessary had the same dollar been earned through expanded exports. 4.60 The overall lack of economic profitability of import substituting activities across most sectors of production also calls into question the overall bias toward import substitution as a criterion guiding both sectoral choice and project choice both in the 1976-80 plan and over the longer term. The DRC calculations indicate that emphasis on import substitution rather than export promotion as a means to earn foreign exchange was an inappropriate guide to investment allocation, at least by the end of the 1976-80 period when substantial import substitution investment had already occurred. Several qualifications to this conclusion should be noted, however, including the following: first, the DRCs for import substitution activities tend to be lower in the priority raw material and energy sectors, with the exception of nonferrous metals, and in the priority agroindustrial complex (agriculture and food processing) than they are in the other industrial sectors. This finding suggests that, assuming import substitution as a general policy objective, the 1976-80 plan identified the "least costly" or "least unprofitable" sectors with which to realize this objective.  - 84 - 4.61 Second, although exporting activities appear socially profitable by DRC calculations in all sectors, external limitations on some Yugoslav exports to the convertible area and/or implicit or explicit costs attached to a major shift of exports to the clearing area might make the profitable expansion of exports in some sectors either more costly than the numbers indicate or impossible. In addition, Yugoslav planners, like planners throughout Eastern Europe and in many other developing countries, assess import substitution activities to be less risky than export activities, since the ultimate performance of the latter depends on world conditions over which they have little or no control. Risk aversion on the part of planners combined with uncertainty in world market conditions would probably have led planners to tolerate somewhat higher DRCs on import substituting activities than on exporting activities. 4.62 Third, because the DRCs are sectoral averages and are calculated ex post, they conceal a wide range of intrasectoral variation. Thus the results do not preclude the possibility that carefully chosen new investment projects in apparently unprofitable activities might nevertheless be profitable. For example, an import substituting project of optimal scale in ferrous metals or chemicals might be productive even though many import substituting projects in these sectors have been unprofitable because of insufficient size. However, to the extent that sectors were chosen as the unit in planning, these results remain pertinent. C. Exchange Rate Policy 4.63 As the foregoing analysis has indicated, the Yugoslav economy was subjected to a series of external and internal disturbances in the 1976-80 period: terms of trade changes, oil price increases, rapid global inflation, increases in nominal and real international interest rates, declining real levels of workers' remittances, deteriorating export performance and a rapid acceleration of domestic inflation. In such an environment decisions on how to move the nominal exchange rate were complex. Maintenance of a given real rate was difficult given the rapidity, scale and heterogeneity of price movements inside Yugoslavia, and among the markets to which it sells. In addition, certain of these developments implied a shift in the real exchange rate wbich was difficult to assess accurately. 4.64 Exchange rate policies can be evaluated both witb respect to their impact on external price competitiveness (in goods where the country is to some degree a price setter), and with respect to their impact on the domestic allocation of resources and hence the supply of exports. The analysis of price competitiveness can be quite sensitive to the particular weighting schemes chosen in aggregating the various markets to which a country exports 1/. Table 4.11 provides information on trends in external 1/ For a full discussion, please see Report 2972-YU Export Performance and Policies, Annex III and references therein.  - 85 - competitiveness based botb on a direct comparison of export unit value indices in dollars, and a comparison of wholesale price-deflated or 'real' effective exchange rate movements. The export unit value comparisons compare Yugoslav exports witb exports from western markets - i.e. they treat Western suppliers as competitors of Yugoslav exporters in tbird markets while the price deflated effective exchange rate indices compare Yugoslav exports with price and cost trends within the developed market economies, implicitly regarding the domestic productive structure within these markets as the source of competition. The market distribution of Yugoslavia's trade creates certain difficulties for these measures, particularly as the export unit value indices reflect all of Yugoslavia's exports, including that which is conducted under bilateral trade agreements where goods traded are covered by commodity lists. Both because the commodity composition of trade differs substantially by the various market areas discussed in Chapter II, and because the price of equivalent goods and price movements over time may differ between these market areas, these aggregate trends represent only an imperfect proxy for the true evolution of export unit values in goods actually exported by Yugoslavia to the developed market economies. Interpreting trends in relative export unit values is thus somewhat ambiguous: they can reflect either the declining competitiveness of Yugoslavia's exports to the markets of the developed market economies, or the fact that prices have tended to rise faster in the markets to which Yugoslavia exports than in the markets of the developed market economies more generally. 4.65 Examining the evidence in Table 4.11, the first point of interest is the relatively little change that occurred in the nominal exchange rate between the dinar and the dollar between 1975 and 1979 - approximately 10%, on a 1975 base, despite the much higher rate of inflation in the dinar than the dollar. An examination of the trade-weighted effective exchange rate index however indicates that movements vis a vis the dollar are a poor guide to the effective nominal depreciation, given the direction of Yugoslavia's trade with the developed market economies. The high weight of the Deutsche Mark (DM) in this trade, and relative shifts between the dollar and the DM in this period result in a trade-weighted depreciation of the dinar of about 17% between 1975 and 1979, rather than the 10% indicated by the dollar rate. It appears, moreover, that when this trade-weighted effective nominal rate is deflated by relative wholesale prices, the price deflated effective rate stayed more or less level over the four year period, appreciating slightly in 1976 and 1977 and then depreciating by an equivalent amount in 1978 and 1979. For the first four years of the plan therefore, the competitiveness of the exchange rate, as measured by the wholesale price deflated series, remained essentially unaltered. A similar result follows from the comparison of relative export unit values in the 1975-79 period. It is only the index deflated by unit labor costs in manufacturing that displays any significant loss of competitiveness in this period. All three indices however show a marked improvement in competitiveness in 1980, both as compared to 1975 and as compared to 1979, following a major dinar devaluation in June 1980.  - 86 - Table-4.11: EXCHANGE RATES AND INDICES OF EXTERNAL COMPETITIVENESS Indices of Price Deflated Effective Trade- -Exchange Rates 4/ Weighted Export Deflated Deflated Period Average Effective Unit Values 3/ by by Relative Exchange Rate Exchange Developed Relative Unit Labor (Dinars Rate Yugo- Market Wholesale Costs in - ---per dollar) 1/ Index 2/ slavia Economies Prices Manufac. 1975 17.39 100 100 100 100 100 1976 18.19 103.6 104 101 104.2 109.5 1977 18.30 100.6 117 109 104.3 111.8 1978 18.64 89.7 127 122 100.5 111.7 1979 19.00 83.2 147 142 100.4 111.9 1980 24.91 66.0 175 160 94.4 91.1 1/ Line rf in IMF, International Financial Statistics. 2/ Weighted by bilateral external trade flows with developed market economies in 1975. 31 US Dollar index from International Financial Statistics. 41 Weighted by exports to developed country trading partners in 1975, representing foreign exchange value of the Yugoslav dinar adjusted for Yugoslavia's prices or costs relative to those trading partners. Source; United Nations, Statistical Yearbook and Monthly Bulletin of Statistics; IMF, International Financial Statistics. 4.66 The evidence thus suggests that, as compared with a 1975 benchmark, there was at most a slight deterioriation in price competitiveness in the period until 1979, though perhaps a somewhat more substantial deterioration in cost competitiveness. The relative stability in the price deflated effective exchange rate is not unexpected, since it was the National Bank of Yugoslavia's explicit objective to move the dinar sufficient to offset movements in relative wholesale prices between Yugoslavia and its trading partners. The evidence in Table 4.11 suggests that the NBY succeeded in  - 87 - achieving this objective. The more relevant and important issue however is whether an excbange rate policy designed merely to offset differential inflation between Yugoslavia and the developed market economies was adequate given the shocks suffered by the economy in this period. The answer to this question is more complicated, and simulation runs of the CGE model have been used to address it. 4.67 The simulation results are based on a model solution which estimates the market clearing exchange rate for each year during the 1976-80 period. These estimates are intended to approximate the exchange rate that would have emerged had the Yugoslavs permitted the exchange rate to be determined by market forces, conditioned by existing tariff and subsidy rates. The estimates assume the elimination of the combined fixprice and flexprice rationing schemes for foreign exchange allocation discussed earlier, and their replacement by exchange rate changes. The estimates are based on Yugoslavia's actual experience in the 1976-80 period, including the actual inflation rates experienced by Yugoslavia and its trading partners and the actual export levels. However, to the extent that capital flows and reserve movements were not "normal" or sustainable, the resulting exchange rate estimates are not realistic measures of an equilibrium exchange rate, since such measures should be based on capital flows and reserve levels that could be sustained over time. While it is not obvious what normal borrowing or reserves behavior would have looked like over the 1976-80 period, some rough assumptions can be made and incorporated in the model simulations to yield other, more realistic, measures of a sustainable market clearing exchange rate. 4.68 It seems reasonable to assume that reserves should have remained sufficient to meet a target of two months import cover each year during the plan period. At the end of 1975, gross foreign exchange reserves stood at $1019 million. After a sharp rise to $2298 million in 1976, they remained above the target of two months import cover through 1978, after which they fell dramatically. By the end of 1980, they amounted to $1462 million and covered less than one month of total imports (equivalent to one and one half months of imports from the convertible currency area). In the simulation results on which the estimates of the equilibrium exchange rates are based, a more normal pattern of reserve accumulation is adopted, in which the ratio of total reserves to total imports is allowed to drop from its above-target levels of 1976-78 to target levels of two-month cover in 1979 and 1980. This pattern implies that reserves could have been reduced in 1979 (although by a much smaller amount than actually occurred) to bring them down to a two-month cover for total imports, although they would have had to increase in 1980, contrary to what actually occurred, to maintain this cover. 4.69 It is more difficult to define a sustainable level of external borrowing because what is sustainable depends in part on bow foreign borrowing affects the productivity, growth and trade performance of the economy and in part on conditions prevailing in international capital markets. While there is no obvious way to assess a sustainable level of borrowing, since the 1976-80 growth of borrowing in both nominal and real terms (as measured by  - 88 - growth in nominal and real medium- and long-term external debt) differed relatively little from what had occurred during the 1971-75 period, and because the debt:GNP and debt service ratios in Yugoslavia remained well within observed ranges for other newly industrializing countries for this period, the model simulation assumes that the historical levels of borrowing were sustainable. 1/ (See Chapter II for a more extensive discussion of capital and reserves flows during the 1976-80 period.) 4.70 With these assumptions on reserves levels and capital flows, the CGE model can be used to estimate values of the exchange rate that would have equilibrated the demand and sustainable supply of foreign exchange. It is important to keep in mind that the exchange rate estimates are estimates of the exchange rate that would have cleared the foreign exchange market given Yugoslavia's actual economic structure and policy regime, including its sectorally differentiated tariffs and direct export subsidies. They are not the same as the estimates of a shadow exchange rate as used in the DRC calculations of Table 4.10, since those estimates additionally involve the removal of price-distorting import tariffs and direct export subsidies. 4.71 Simulations of the CGE model on which estimates of the market clearing exchange rate are based are called flexible exchange rate runs. These results are best understood in comparison with the results of the historical base run of the model. In the historical base run, the exchange rate is fixed at period average, official parity levels in each year, and the model solution is based on the export levels, capital flows and reserves changes that actually occurred. Since, through the entire 1976-80 period, there was a shortage of foreign exchange at the official exchange rate (see Table 4.7), the historical run uses the fixprice and flexprice rationing mechanisms described earlier to allocate foreign exchange. By contrast in the flexible exchange rate run the model finds a market clearing exchange rate level that eliminates the need for these rationing mechanisms. Such estimates provide a reference path of exchange rates that would have emerged given the plan strategy, the policy framework and the exogenous shocks that actually characterized the 1976-80 period but assuming a flexible exchange rate policy and "normal" reserve behavior had been pursued. 4.72 The resulting estimates of market clearing exchange rates are presented in Table 4.12. A comparison between these rates and the official exchange rates indicates that even in 1976, the first year of the plan, there was significant overvaluation of the official exchange rate, of approximately 23%. This degree of overvaluation lessened in 1977 reflecting the liberali- 1/ In the light of developments in 1980 and 1981 this assumption may be somewhat optimistic.  - 89 - zation of imports that occurred in that year but increased thereafter through 1979, and remained high in 1980 despite the change in the nominal exchange rate that occurred in that year. These estimates should be regarded as upper bounds of the degree of overvaluation because it maybe unrealistic to think of the elimination of all quantitative restrictions on imports in measuring the market clearing rate. The evidence nevertheless does suggest that in terms of the balance between the demand and the supply of foreign exchange on the domestic market there was significant, persistent and growing overvaluation of the official exchange rate of the dinar over the period, an overvaluation which necessitated the growing use of the various fixprice and flexprice rationing mechanisms described earlier; and which considerably aggravated the difficulties of implementing the new foreign exchange allocation system. Table 4.12: ACTUAL AND MARKET CLEARING EXCHANGE RATES, 1976-1980 (Dinars per US Dollar, Period Average Estimates) Ratio of Market Actual Market Clearing Clearing to Actual Exchange Rate a/ Exchange Rate b/ Rate (x) 1976 18.19 22.33 1.23 1977 18.30 20.20 1.10 1978 18.64 22.55 1.21 1979 19.00 26.91 1.42 1980 24.91 33.47 1.34 a/ Line rf of Yugoslavia in International Financial Statistics. b/ Based on CGE model for Yugoslavia. 4.73 Quite apart from differential inflation between Yugoslavia and its trading partners, other developments in the 1976-80 period had an effect on the demand and supply of foreign exchange. The most important of these were perhaps the reduction in the growth rate of exports discussed earlier, the reduction in net real worker remittance flows, and the rise in the price of oil in 1979-80; between them these factors were the major elements in the foreign exchange crisis of 1979-80. The CGE model constitutes a tool to examine the relative importance of each of these developments. It does so by providing an answer to the question: what would the market clearing exchange rate in 1980 have been if one or all of these elements had behaved significantly differently over the period. 1/ This answer is provided by comparing the 1980 market clearing exchange rate derived from simulations of the historical flexible exchange rate run, with estimates of the 1980 market clearing rate generated by runs of the model which substitute other, counterfactual assumptions instead of the actual historical experience. 1/ See Appendix I for details of these experiments.  - 90 - 4.74 A summary description of each of these simulations and the 1980 market clearing exchange rate that results for each is provided in Table 4.13. In the 'historical' flexible exchange rate simulation, the dollar values of exports are set at their actual, realized levels for each year. In simulation B-1 export growth is set at its trend level. This entails raising the overall growth rate of industrial merchandise exports from an estimated actual value of 2.9% for the plan period to its 1965-75 trend value of 6.2% and raising the actual growth rate of total nonfactor service exports from an estimated actual value of 7.4% to its 1965-75 trend value of 12.0%. Since the actual growth rate of agricultural exports at 6.4% during the 1976-80 period did not differ much from the 1965-75 trend rate, the actual rate was left unchanged in simulation B-1, although as in the case of both industrial and service exports, the experiment specified constant annual export growth rates at trend values for each year, thereby smoothing out the erratic fluctuations in export growth that actually occurred during the period. 4.75 Simulation B-2 incorporates the assumptions of simulation B-1, but further sets the domestic inflation rate at the weighted average inflation rate experienced by Yugoslavia's trading partners. This experiment is designed to estimate the effects of the "pure" inflation differential on the changc in the equilibrium exchange rate. While the other factors analyzed in the simulations provide estimates of changes in 'real' exchange rates required to balance demand and supply in the domestic market, this experiment provides an estimate of the change in the nominal rate required to maintain a given real rate. According to the 1976-80 data used in the model, Yugoslavia's annual inflation rate was an estimated 19.0%, while the "trade inflation rate" reflected in a weighted average of its import and export prices was only 13.4%, leading to an inflation differential of 5.6%. Even in the absence of any other internal and external shocks, the existence of such a differential would have necessitated some nominal depreciation of the dinar, in order to neutralize the effects of this inflation differential on the structure of incentives in the domestic economy. Simulation B-2 is designed to estimate the depreciation that would have been necessary to achieve this objective. 4.76 Simulation B-3 examines the effects of the 1979-80 oil price increase by examining what would have happened to the 1980 equilibrium exchange rate if the price of imported oil had increased at the same rate as all import prices in 1979 (15%) and 1980 (18%), rather than at the actual estimated rates of 45.3% and 78.5% for those years. Finally, simulation B-4 looks at the effects of the slowdown in net remittances that occurred over the 1976-80 period, at least partly in response to external conditions that led to a steady reduction in the number of Yugoslav workers employed abroad. The experiment assumes that net remittances in nominal  - 91 - Table 4.13: ALTERNATIVE ESTIMATES OF MARKET CLEARING EXCHANGE RATES, 1980 Estimated 1980 Simulation Description Exchange Rate Historical simulation Fixed excbange rate; historical 24.91 values of parameters and exogenous variables. Foreign exchange market cleared by rationing mechanisms Flexible Excbange Rate Simulations Historical flexible Reserves set at two-month 33.47 exchange rate simulation cover for imports in 1979 and 1980; otherwise, parameters and variables at historical values. Foreign exchange market cleared by exchange rate adjustment. B-1 Export growth at trend rates 29.53 B-2 B-1 plus domestic inflation set 24.37 at trade-weighted inflation rates of trading partners. B-3 B-2 plus no oil price shock in 23.01 1979-80 (oil prices growing with average import prices). B-4 B-3 plus net remittances constant 21.44 in real terms. terms grew in line with a weighted average of Yugoslavia's import and export prices, leaving real net remittances constant. The assumption implies a cumulative difference of