Angola - Trade, price, and wage reform

Repor No. 8964-ANG!- ..' ' Angola Trade, Price, and Wage Reform June 20, MI9 C,ountry Operations Division South Central and Indian Ocean Department Africa Relgion FOR OFFICIAL USE ONLY _ _* Ci--A;--;; -- ' :YS;-*'%S'#o* ,'.. tt < ',:~~~~~~~~~~~~2  CURRENCY EQUIVALENT (as of March 31, 1991) Currency Unit - NoW Iwanza Official Rate: US$1 = NKz 60.00 FISCAL YEAR January 1 - December 31  FOR OFFCIAL WI ONLY ANGOLA TRADE, PRIC AND WAGE REFORM County Operatla ivsOM" ouhCmin"a ad ndian Oean Department Aftm ROOiM This rport i basd on te fidins of an economic mission tat visised Angola in November/D ber 1990. Mision members wele Cais Elbirt (mission eade), Hecwt Avi4, Mawa Taa Benio. Roa Castro, Per Pall, Daie Kaufa, and Lob Ptere da Silva. With a tion of Hecor Avila (con_aa), all are Wod Bank staff membem Th fepo wa dicusd with th Ansolat in Februay 1991. Thi docuent hw a rsid distributio and may be ud by rcpients only In the pfome of their cal dutisu Its contents may not othwise be lod withot Worldnk authorition.  COUUTD DAT& .*.......... ..... .... .... ...... vij WMt0UH* 8T Y .*........................... ix CRWTR I TJRDAID ERGM ........ ... .. ... .. ..... . 1 I. A SIMPLIFIED ANALYSIS OF THE TRADE REGIME IN AFRICA . . . . . . . . 1 II. THE ANGOLAN TRADE REGIME . . . . . . . . . . . . . . . . . . . . . . 3 A. External Trade . . . . . . . . . . . . . . . . . . . . . . . . . 4 1. Foreign Exchange Allocation System . . . . . . . . . . . . . 4 2. Structure of Imports . . . . . . . . . . . . . . . . . . . . 7 3. The Parallel Market for Foreign Exchange . . . . . . . . . 10 S. Internal Trade ..10 1. Price Controls . . . . . . . . . . . . . . . . . . . . . . 11 2. The Parallel Market for Goods . . . . . . . . . . . . . . 13 C. Other Trade Policies . . . . . . . . . . . . . . . . . . . . . 15 1. Import Tariffs and Sales Tax . . . . . . . . . . . . . . . 15 2. Export Restrictions . . . . . . . . . . . . . . . . . . . 16 III. IMPLICATIONS OF THE CURRENT TRADE REGIME . . . . . . . . . . . . . 17 A. Costs Associated with Market Distortions . . . . . . . . . . . 17 B. Sectoral Impacts of the Trade Regime . . . . . . . . . . . . . 17 IV. THE POLICY IMPLICATIONS AND SUGGESTED OPTIONS FOR REFORM . . . . . 18 A. Exchange Rate Unification . . . . . . . . . . . . . . . . . . 18 B. Integration of the Internal Market . . . . . . . . . . . . . . 19 C. Complementary Measures ........ .. ... .. .. .. . 20 ANNEX 1 DETERMINING PRICES IN THE PARALLEL MARXET . . . . . . . . . . . 21 A. General Conclusions . . . . . . . . . . . . . . . . . . . . 21 B. "Nontradable" Imported Fixed-Supply Goods . . . . . . . . . 21 C. Tradables ......................... . 22 QWMAM 2 PUXC TIMM= . . . . . .... ......... .. .. . 24 I. AN OVERVIEW OF THE PUBLIC SECTOR . . . . . . . . . . . . . . . . . 24 II. THE BUDGETARY ACCOUNTS . . . . . . . . . . . . . . . . . . . . . . 25 III. MEASUREMENT OF BUDGETARY OPERATIoNS . . . . . . . . . . . . . . . 25 IV. THE BUDGET DEFICIT, EXCHANGE RATE, AND PRICES . . . . . . . . . . 28 V. a CONSOLIDATED ACCOUNT FOR THE 13 LARGEST PUBLIC ENTERPRISES . 32 A. Public Enterprises in Angola . . . . . . . . . . . . . . . . 32 B. A Consolidated Account for the 13 Largest Parastatals . .. 34 1. Production, Employment, and Investment . . . . . . . . . . 35 2. Revenues, Expenditures, and Transfers . . . . .35 C. Measuring Real Resources Mobilization for the 13 Largest Parastatals . . . . . . .37 D. Effects of Policy Changes on Parastatal Accounts. 3-  - iv - ANNEX 2.A INSTITUTIONAL ASPECTS AND DESCRIPTION OF THE BUDGET . . . . . 43 A. Budget Formulation, Execution, and Control . . . . . . . . . . 43 B. Revenues, Expenditures, and Financing . . . . . . . . . . . . 44 1. Revenues . . . . . . . . . . . . . . . . . . . . . . . . . 44 2. Expenditures .... . . . . . . . . . . . . . . . . . . . 45 3. Financing . . . . . . . . . . . . . . . . . . . . . . . . 46 C. Extrabudgetary Operations ... . . . . ... . . . . . . . . 47 ANNEX 2.B ANJ "ADJUSTED" BUDGET AND THE EXCHANGE RATE . . . . . . . . . . 48 ANNEX 2.C THE GOVERNMENT BUDGET AND THE EXTERNAL ACCOUNTS . . . . . . . 49 ANNEX 2.D THE INSTITUTIONAL RELATIONS BETWEEN THE CENTRAL GOVERNMENT AND THE PUBLIC ENTERPRISES . . . . . . . . . . . 51 ANNEX 2.E METHODOLOGICAL ISSUES REGARDING THE CONSOLIDATED PARASTATAL ACCOUNTS ... . . . . . . . . . . . . . . . . . 52 CHAPTUR 3 RZMONZTIXASIO= OF INCOMES IN TBH FORMAL SCIOR . . . . . . . . 53 I. LABOR FORCE AND THE FORMAL SECTOR IN ANGOLA. . . . . . . . . . . 54 A. The Formal Sector Labor Force ... . . . . . . . . . . . . . 54 B. The Remuneration System . . . . . . . . . . . . . . . . . . 56 C. Economic Consequences of the Present System . . . . . . . . . 59 II. REMONETIZING INCOMES . . . . . . . . . . . . . . . . . . . . . . 60 A. The Economic Advantages of Remonetization . . . . . . . . . . 60 B. Wage Remonetization Parameters ... . . . . . . . . . . . . . 61 C. The Estimates ... . . . . . . . . . . . . . . . . . . . . . 62 III. THE POLICY IMPLICATIONS OF REMONETIZATION . . . . . . . . . . . . 63 A. Impact of Remonetization on the Public Sector Fiscal Deficit: A Simple Accounting Framework . . . . . . . . . . 64 B. Preparing for Remonetization ... . . . . . ..... . . . . 66 C. Other Possible Policy Implications of the Remonetization . . . 69 D. A Suggested Sequencing for Remonetization . . . . . . . . . . 70 1. A Two-Stage, Top-Down Approach . . . . . . . . . . . . . . 70 2. Estimates of the Financial Shift from Remonetizing Two Entitlements . . . . . . . . . . . . . . . . . . . 71 E. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 71 ANNEX 3 AN ACCOUNTING FRAMEWORK FOR THE ANGOLAN ECONOMY AND THE REMONETIZATION ... . . . . . . . . . . . . . . . . . . 74 A. The Model ... . . . . . . . . . . . . . . . . . . . . . . . 74 B. Monetization of This Economy ... . . . . . . . . . . . . . . 75 T!SPLS 1.1 Key Economic Indicators in Selected Sub-Saharan African Countries, 1982 . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Matrix for Analysis of the Angolan Foreign Exchange Allocation System ....................... . 5 1.3 Per Capita GDP and Imports in Selected African Countries, 1987 and 1988 ......................... . 8 1.4 Official Foreign Exchange Allocation for Noncapital Imports, 1985-90 ........................ . 9 1.5 Parallel Market Exchange Rate in Luanda, 1984-89 . . . . . . . . . 11 1.6 Official Retail Prices and World Prices for Selected Goods, 1980, 1985, and 1989 .................. . 14  - v - 1.7 Parallel Market Prices for Selected Products, Luanda, 1985and1989 .... . . . . . . . . . . . . . . . . . . . . . 16 2.1 Central Government Finances, 1967-89 . . . . . . . . . . . . . . . 26 2.2 1989 Budget Including Extrabudgetary Expenditures . . . . . . . . 27 2.3 Foreign Exchange Component of Central Government Expenditures, 1989 Budget . . . . . . . . . . . . . . . . . . . 28 2.4 Government Revenues at Parallel Prices, 1989 . . . . . . . . . . . 29 2.5 Government Expenditures at Parallel Prices, 1989 . . . . . . . . . 30 2.6 Adjusted Government Deficit, 1989 ..31 2.7 Structure of "Adjusted" and "Nonadjusted" Budgets, 1989 . . . . . 31 2.8 Enterprises by Size and Type of Ownership, 1989/90 . . . . . . . . 33 2.9 The 13 Largest Public Enterprises, 1989 . . . . . . . . . . . . . 34 2.10 Consolidated Nonadjusted Account for 13 Parastatals, 1986-90 . . . 38 2.11 Consolidated Nonadjusted Account for 13 Parastatals by Economic Sector, 1986-90 . . . . . . . . . . . . . . . . . . 39 2.12 Consolidated Nonadjusted Account for 13 Parastatals by Economic Sector, 1986-90 . . . . . . . . . . . . . . . . . 40 2.13 Consolidated Nonadjusted Account for 13 Parastatals, 1986-90 . . . 41 2.14 A Simulated Consolidated Account for the 13 Largest Parastatals, 1986-90 . . . . . . . . . . . . . . . . . . . . . . 42 2.15 Central Government Revenues, 1987-89 . . . . . . . . . . . . . . . 44 2.16 Central Government Expenditures, 1987-89 . . . . . . . . . . . . . 46 2.17 Central Government Deficits and Financing, 1987-89 . . . . . . . . 47 2.18 Dollar Component of the Government Budget, 1989 . . . . . . . . . 49 3.1 Total Population and Economically Active Population in the Formal and Informal Sectors, Selected Years, 1980-90 . . . . 55 3.2 Sectoral Breakdown of the Labor Force in Angola, Mozambique, and Central Africa . . . . . . . . . . . . . . . . . 56 3.3 Breakdown of the Formal Sector Labor Force by Occupational Category, 1988-89 ..57 3.4 Valuation of Basic Entitlements at Official and Parallel M-arket Prices, 1984 ..63 3.5 Remonetized Monthly Wages of Formal Sector Workers by Occupational Category, 1989 ..64 3.6 Remonetized Monthly and Annual Wage Bill in the Formal Sector by Occupational Category, 1989 . . . . . . . . . . 65 3.7 Total Annual Wage Bill in 1989 Including the Military Under Both Systems of Remuneration ..66 3.8 Structure of Wage-Related Expenditures in the Central Government Budget, 1989 ..67 3.9 Suggested Sequencing for the Remonetization of Managers' and Technicians' Salaries ..72 3.10 Examples of Salary Increases for Managers and Technicians . . . . 73 3.11 The Financial Implications of a Partial Remonetization for Managers and Technicians .73 Statistical Appendix Tables 1. Retail Official and Parallel Market Prices, Luanda . . . . . . . . 79 2. Composition of Importse 1985-1988 ..80 3. Taxes on Petroleum Sector ..81 4. Central Government Employment and Salaries, 1988 . . . . . . . . . 84 S. Foreign Exchange Component of Central Government Expenditures . . 85 6. Debt Service as of 31/12/88 ..86 7. Balance of Payments, 1987-90 ..87 8. Population, Economically Active Population, Formal and Informal Sector . . . . . . . . . . . . . . . . . . . . . . 88 9. Breakdown of the Labor Force by Economic Sector, 1988 . . . . . . 89 10. Formal Sector Civilian Employment, by Region and by Sex . . . . . 90  - vi - 11. Population, 1900-86 . . . .*. 91 12. Economically Active Population by Economic Sector, 1985 .i . . . . 92 Annex Thirteen State Enterprise Accounts ... . . . . . . . . . . . . . 93 A.1 General Case: Market Equilibrium under Import Rationing . . . . . 21 A.2 Market Equilibrium with Diversions from Official to Parallel Markets . . . . . . . . . . . . . . . . . . . . . . 22 A.3 Market Equilibrium with Liberal Windows . . . . . . . . . . . . . 23 2.1 Volume of Production, 1980-89 ... . . . . ... . . . . . . . . 35 2.2 Total Parastatal Revenue by Sector, 1986-89 . . . . . . . . . . . 35 2.3 Total Parastatal Expenditures by Sector, 1986-89 . . . . . . . . . 36 2.4 Net Transfers from Parastatals to the Government, 1986-89 . . 36 3.1 Sectoral Breakdown of the L&!'or Force, 1988 . . . . . . . . . . . 56 3.3 Occupational Breakdown in the Formal Sector, 1988 . . . . . . . . 57 3.3 Components of Employee Remuneration, 1988-89 . . . . . . . . . . . 59 3.4 Monthly Income by Occupational Category, 1988-89 . . . . . . . . . 61 1. Angola  - vii - roOUffR DM Anoa Area (thousasnd of square kiloMute.) 1,276.7 population (1989) (million) 9.7 Rate of growth since 1980 (percent) 2.7 Population density (1986) (per square kilometer) 7.0 Population characteristics (1980-85) Crude birth rate (per 1,000) 47 Crude death rate 'per 1,000) 22 Inaonm distribution Percent of national income Highest quintile Lowest quintile Nutrition (1980) Calorie intake as percent of requirements 87 Per capita protein intake (grams/day) Health (1986) Population per physician 15,521 Population per hospital bed 634 Infant mortality (per 1,000 live births) 160 Access to safe watex (1985) (percent of population) Urban 80 Rural 15 Access to sewerage (1985) (percent of population) Urban 26 Rural 16 HneWgy consumption per capita (1986) (kilograms of oil equivalent) 299 Education (1985) Adult literacy rate (percent) 28 Primary school enrollment (percent) 44 Gross dometic product (1989 in US$ millions) GDP at market prices 7.4 MNP per capita (US$) 620 Investment Gross domestic savings Resource balance 735 Exports of goods and nonfactor services 3,151 Imports of goods and nonfactor services 2,416  - viii - Costrl gove et f5iaaa 1969 percent (billions of kwanzas) 1987 1988 budget of =1 (1969) Current revenue 62.9 73.3 81.1 36.5 Current expenditure 76.4 79.2 89.1 48.1 Current surplus -13.5 0.1 -8.0 -3.6 Capital expenditure 11.0 16.4 20.1 9.1 Money and credlt (billions of kwanzas) 1984 1985 1986 1987 1988 Money and quasi-money 172 204 220 257 313 Bank credit to central government 118 130 187 247 297 Other bank credit 97 113 78 54 53 Money as percent of GDP 89.0 101.0 112.5 113.6 Annual percentage changes in Bank credit to central government 48.7 10.1 43.6 32.2 20.0 Other bank credit -10.9 15.2 -30.7 -30.4 -2.9 Balance of payment. 1984 1985 1986 1987 1988 1989 Exports (goods and nonfactor serviceo) 2,129 2,344 1,406 2,409 2,606 3,151 Imports (goods and nonfactor services) 2,058 1,928 1,543 2,159 2,181 2,416 Resource balance 71 416 (137) 614 36 735 Net factor income (314) (329) (370) (218) 538 -751 official transfers (net) 26 20 139 52 32 -4 Current account balance (217) 107 (368) 448 -470 -20 Public medium- and long- term borrowing (net) 173 171 7 (55) -199 -155 Other 101 (260) (283) -834 -256 -754 Change in reserves (- increase) (57) (18) 78 -17 -49 6 Gross official reserves (end of period) 240 257 279 190 239 233 major m An ezpots (average 1984-88) U#$ nllion vercet Crude oil 1,601 87.4 Refined oil 73 3.5 Liquefied petroleum gas 27 1.3 Diamonds 85 4.1 Coffee 52 2.5 other 23 1.1 Total 2,062 100.0 Uzternal debt, Decombes 31, 1989 Total outstanding and disbursed (US$ billions) 6.3 Debt sevice ratio (1987) Total outstanding and disbursed (percent) 21.9 .. Indicates that data are not available.  ZXKCUTZEr su_am 2.. This report complements an earlier report prepared by the Bank as Executing Agency for the U.N. Development Pro'wram (UNDP) {"Angola: An Introductory Economic Review," No. 7408-ANG, June 1590). It is intended as an analytical and policy "tool-kit" for use by the government in its ongoing reflection on viable economic policy refore options. Accordingly, it emphasizes the "how" as well as the "what, " through the application of detailed methodology and several technical annexes (which are appended to each chapter for convenience). However, it is important to stress that not all analytical findings and methodological suggestions necessarily imply policy recommendations. As repeatedly stated in the report, the key polioy reforms should not be taken in isolation, but as part of a coherent policy package. Policy reform is especially important and has become more feasible as a result of the ce&sefire and the signing of the Bitoril poace accords on May 31. 2. Angola is the third largest country south of the Sahara. It has a population of about 10 million, growing at a rate of about 2.7 percent a year, and a low density of 7.2 inhabitants per square kiloeter. The country is very rich in natural resources (especially oil and diamonds) and arable land. However, Angola has suffered from armed confliet for decades, first in the struggle for independence and subsequently in a long and destructive civil war. 3. The development of the Angolan economy accelerated considerably after World War SI. It was stimulated initially by a coffee boom, which made Angola one o0 the world' a four main coffee exporters by 1974. The most spactacular development was that of oil production, which started in the late 1950s and reached 144,000 barrels a day in 1973. The favorable economic opportunities attracted thousands of Portuguese settlers, whose number increased from 40,000 in 1940 to 340,000 in 1974. A large propo-tion of the rural population was mployed in the plantations, mines, and factories either voluntarily (at very low wages) or under a system of forced labor, which was abolished only in 1961. Despite positive economic developments, the vast majority of Angolans continued to live in poverty and social indicators remained extremely low. 4. Disruptions created by the armed struggle and the exodus of the Portuguese settlers during the transition to independence in 1975-76 led to a decline in productive activities. The oil industry, an economic enclave, and diamond mining have been the only important exceptions to the general economic decline. By early 1991, oil output had reached about 460,000 barrels a day. Oil export revenues, comprising over 90 percent of total export revenues, and equivalent to about US$300 per capita, provide the foreign currency needed for imports of war materiel, production inputs, food for the urban population, and other consumer goods. 5. The poor performance of the nonoil economy is explained largely by three factors. First, the war has created insecurity, destroyed economic and social infrastructure, imposed a heavy burden on the economy, and created enormous suffering and deprivation among the population. The second factor is the chaos created by the massive exodus at ndependence of about 300,000 Portuguese settlers (90 perc)nt of the total) who held practically all skilled jobs. Despite the progress made by Angola in education, the scarcity of trained personnel continues to be a major constraint to economic development. The third factor - and the subject matter of this report - is the inappropriate economic policies, which have contributed significantly not only to the poor performance of the nonoil economy, but also to inequalities in income distribution. Without the contribution of expanding oil production, the economic crisis in Angola would  - x - have been even more serious. The resulting dualistic development of the economy has been reinforced by rigid economic management and policies. 6. In May 1991, the government and UNITA (the main opposition force) agreed to a ceasefire as a preliminary step to an overall political settlement and multiparty elections in 1992. Thus, for the first time in decades, the prospects for an end to war and for a permanent political settlement of the Angolan conflict are good. This will permit the country to move toward rehabilitation and economic recovery. At the same time, it increases the importance of improvements in economic policy. The ecomsoic Policy P:oblei 7. The economic policies implemented since 1975 have not been conducive to development. The authorities maintained expansionary fiscal and monetary policies while freezing producer and consumer prices, nominal wages, and the exchange rate at artificially low levels. This created large and mounting external and internal imbalances. The price and wage freeze has encouraged the emergence and steady growth of parallel markets and has constituted a strong incentive to divert resources from the official to the parallel economy. In an attempt to protect the real incomes of formal sector workers, the government introduced a system of purchasing rights at official prices, which constitutes a substantial part of wages. The large currency overvaluation has resulted in a virtual elimination of all exports except oil and diamonds, a heavy dependence on imports, and an ^conomy in which rent-seeking probably constitutes the most important activity in urban areas. The economy of Angola is characterized by extreme price distortions across the board. The official exchange rate is one one-hundredth of the parallel one, and parallel prices are 20 to 100 times official prices. 8. Recognizing that Angola's economic distortions and imbalances are due in large part to inappropriate economic policies, the government announced a program of Economic and Financial Restructuring (SEF) in 1987. The proposed reforms included (1) a limited adjustment of the exchange rate, which would still have left a wide margin between the official and the parallel exchange rates; (2) a reduction in the budget deficit and a corresponding reduction in inflationary financing; (3) a program for restructuring public enterprises through divestiture and liquidittions; (4) reform of the financial sector by encouraging the establishment of private commercial banks and ensuring greater availability of credit to the private sentor; and (5) a cutback in price controls to selected essential goods only. 9. While the proposed reforms were in the right direction, the measures generally did not go far enough to address the main structural distortions, notably in the area of the exchange rate. In any event, with the exception of a more accommodating attitude toward parallel market activities, there has been little progress to date in implementing the program. Since 1988, several laws have been adopted to provide a broad framework for structural reform, but most of them have not been applied. A new package of measures, including a 100- percent devaluation combined with wage increases and price liberalization, was announced in early 1990 as part of a New Action Plan. The exchange rate adjustment, however, was postponed for over a year and the decontrol of most prices is yet to be implemented. 10. A currency reform in September 1990 (creation of a new currency and tight limits on convertibility from the old currency) temporarily brought dcwn prices on the parallel market. However, this attempt to reduce excess liquidity proved unsustainable. On the contrary, the experience of this "reform" further undermined the already shaky public confidence in the government's management of the economy. In early 1991, inflation was accelerating and the gap betwsen official and parallel exchange rates again widening to almost the range that existed before the currency reform. Yet another attempt to put forth a more consistent and comprehensive set of stabilization measures was made recently, as part of the annual plan for 1991. While reaffirming the government's commitment  - xi - to adjustment and reform, this plan still does not represent the kind of comprehensive and credible package of reforms needed to address the fundamental economic distortions of the Angolan economy. For example, the 100-percent devsluation implemented in March 1991 is insufficient to address the substantial overvaluation of the new kwanza. The Overall Peramective 11. The complexity of Angola''s political economy and the degree of distortion make this economy an unusual case for reform. Although the general menu of policy changes is well-known (trade liberalization, devaluation, price liberalization), and the associated fiscal and monetary policies are theoretically well-established, the scope and pace of change in this poor, war- ravaged economy remain to be ascertained. On the one hand, the prospects for economic growth are substantial, given the rapid supply response that one can expect in some sectors (agriculture, services); on the other, the inertia resultinq from both the lack of management experience and the vested interests associated with the current distortions is a considerable obstacle to rapid, radical change. 12. Three crucial policy areas need to be addressed: trade and pricing, public finance, and wage remonetization. Concerning the first, the "Introductory Review Report" made clear that fundamental price distortions in Angola are reflected in the extreme segmentation of markets for goods and foreign exchange. Knowledge of the foreign exchange allocation regime and the relationship between official and parallel markets is still insufficient. Concerning public finance, the need for a better understanding of the public sector was considered crucial, given the preponderant role of the government in the economy and in any prospective adjustment, *.he importance of the parastatal sector, and the distorted and unreliable statistics. In view of the magnitude of the task and the scarcity of data, the report focuses on the central government and some selected public enterprises, emphasizing measurement problems in a highly distorted economy as a starting point for formulating policy prescriptions. Finally, the importance of wage remonetization is underlined by the fact that the official economy is governed basically by coupons and buying rights, with the local currency playing a minor role. The need to understand this system i3 overriding. 13. This report's focus on economic policy reform does not mean, of course, that Angola's development will not also require a stable and peaceful climate, rehabilitation of infrastructure, and institutional capacity building. On the contrary, these are intertwined, and sustainable development will require all of them. Correction of price distortions will principally involve movements in the exchange rate, price liberalization, and a concomitant remonetization of wages. These measures will have to be complemented by sound fiscal and monetary policies. The 1ocus of AMIstuUnt 14. As noted, the central focus of policy adjustment in Angola should be the elimination of price distortions. This would automatically improve the utilization of oil revenues and could generate a significant supply response in the tradables sector. Economic distortions in Angola have had a negative impact on resource allocation and income distribution. Policy reform should therefore aim at market integration (rejoining official and parallel markets), economic restructuring (rebuilding the nonoil economy), macroeconomic stability (controlling inflation, avoiding sharp fluctuations in incomes), and the protection of vulnerable groups (maintaining some protection for formal sector workers through entitlements, implementing other measures to protect informal sector workers). Within this framework, a reform package will have to include radical adjustment of the exchange rate, prices, and wages. This will inevitably have budgetary and monetary implications.  - xii - 15. Given the degree of currency overvaluation, there is no escape from a major exchange rate adjustment, which should be directed at the eventual unification of the foreign exchange markets. This would allow the introduction of an open generalized licensing (OGL0 system for imports, thus eliminating the administrative allocation of import licenses and foreign exchange. A competitive exchange rate is also crucial to provide adequate incentives for nontraditional exports, particularly in agriculture. An OGL system would improve access to inputs by productive sectors. In addition, the exchange rate adjustment would substantially reduce rent-seeking activities and corruption. 16. Price liberalization and the removal of controls also will be crucial for market integration. This process would need to be accompanied by the introduction of competition in the distribution network. In Angola, price liberalization need not be inflationary; provided appropriate fiscal and monetary policies are in place, prices may even fall. If a decline in real wages is warranted by macroeconomic conditions, it could be implemented simultaneously with wage remonetization. In the course of implementing a wage remonetization, it may be desirable to maintain some basic buying rights (cartaos) for a time for the poorest sectors of the society, to provide a safety net. The Pace of Mdustasnt 17. The process of economic adjustment would require the adoption of a wide- ranging set of measures at the institutional, macroeconomic, and sectoral levels. The macroeconomic adjustments (prices, exchange rates, wage remonetization, fiscal and monetary policy) must provide an appropriate incentive framework for implementing sectoral reforms and should therefore proceed rapidly. Sectoral and institutional changes (public enterprise reform, privatization of public sector assets, financial sector reform) would take longer in most cases. Although this report focuses largely on prices and wages, this should not be interpreted as minimizing the crucial importance of other sectoral and institutional changes, some of which are currently being addressed by the authorities. 18. Some measures should proceed more rapidly than others, but it is still essential to take a global, synchronized approach. For example, institutional changes such as the redefinition of property rights should accompany macroeconomic adjustments if the expected economic response is to materialize. Since comprehensive and radical changes such as those required by Angola demand concerted societal efforts, the pace of adjustment will be affected by political and social factors and by institutional capacity. Measures that consolidate political support, minimize transition costs, and economize on scarce management skills should clearly be taken as rapidly as possible; measures that require consensus building, have high transition costs, or require institutional support, may be paced accordingly. 19. The macroeconomic adjustment could conceivably begin with gradual steps, to "test the waters" and gather more information about the direction and response of the economy to policy adjustments. once there is more confidence in the manageability of the process and the differential between the official and parallel markets has been narrowed, the authorities could decide to complete the adjustment rapidly, with sharp changes in prices, including the exchange rate and wages. 20. Experience has proven that the main phase of adjustment should proceed decisively and be completed in a relatively short period of time; otherwise, initial gains might be eroded. Partial adjustments usually require special incentives to establish, in advance, the general trend intended for the economy. For example, to encourage exports and to promote efficient import-substitution, a competitive foreign exchange "window" or other market-type mechanism could be introduced. However, such partial adjustments could present some complications that a faster pace would not. Enterprises with access to a window should probably operate with fully liberalized prices and remonetized wages, as if they were part of the parallel, or fully liberalized, economy. Furthermore, the links between enterprises operating under a fully liberalized price regime and those  - xiii - continuing to operate under official prices might be disrupted. These disruptions will be minimized, however, if rapid and continued movements take place to narrow the gap between the official and parallel markets. The danger of miscalculating wage increases is less intense in this type of program. Similarly, the monetary implications of wage remonetization appear more manageable. Various options associated with the sequencing and pace of wage remonetization are discussed in more detail below. In any case, whether simultaneous or sequential, rapid or gradual, effective adjustment requires e.;plicit consideration of the linkages between different policy measures. the Key Policy likacaes 21. The following general possible scheme emerges from the linkages among the key variables. An immediate adjustment of the official exchange rate to a market-clearing level (through, say, an auction system or a managed exchange rate) will have to be accompanied by generalized price liberalization. Simultaneously, wages should be remonetized and, if required by the macroeconomic situation, reduced in real terms. An adjustment path that corrected prices, exchange rate, and wages at the same time would rapidly improve resource allocation. However, it would require simultaneous adjustments in prices and wages of a magnitude that could be difficult to manage. Given the extent of the distortions, there is a cLanger of miscalculating wage adjustments in the process of remonetization, thereby generating excess demand and inflationary pressures. A major uncertainty is the level at which liberalized prices will settle, since that will determine the parameters for wage remonetization. The response of public enterprises to the new measures is also a major issue since these enterpri3es are initially the main economic agents for the transmission of such liberalization throughout the economy. Finally, for the program to be successful, these drastic adjustments in prices will need to be accompanied by appropriate monetary and fiscal policies. Credit policy and the level of the monetary aggregates should be consistent with wage policy, including the effects of remonetization, and with the inflation and balance of payments objectives. 22. Consequently, among the variety of linkages between policy variables, three key linkages are most important in Angola: between the exchange rate and prices, between prices and wages, and between wages and monetary policy. 23. The exchange rate-price linkage: In Angola, exchange rate liberalization requires price liberalization, but the opposite is not necessarily true. It is possible to proceed more expeditiously with liberalizing prices than the exchange rate system. However, if the exchange rate is fully liberalized, prices should be liberalized and wages should be totally remonetized as well. 24. The price-wage linkage: Wage remonetization should proceed in line with price liberalization. If prices are totally liberalized, full wage remonetiza- tion should be implemented. If prices are not fully liberalized, wages should be adjusted only to compensate for movements in official prices, and wage remonetization will also be taking place gradually. When official and parallel prices are unified, purchasing rights become irrelevant, and remonetization has been achieved. In a gradual approach, there are other options for remonetizing wages. 25. The wage-monetary policy linkage: A monetary policy consistent with wage remonetization is yet to be determined. The key empirical issue here is the extent to which the replacement of purchasing rights at official prices by money at market prices will lead to an expansion in money supply and what further impact this might have on the market price level. As these measures would be implemented in a coherent context of far-reaching adjustment, the overall economic climate and confidence in the currency would be positively affected. Hence, money velocity would most probably slow down (as it tends to do in countries undertaking a serious adjustment effort), and the inflationary impact of wage remonetization is difficult to predict.  - xiv - !he Main Findihas and golIlc Recommendations 26. Trade and Prices. The mechanism of foreign exchange allocation constitutes the key determinant of the economic situation in Angola. A notable feature of the trade regime in Angola (as in other countries with overvalued exchange rates) is that the foreign exchange allocation system has allowed relatively liberal access to foreign exchange at the official exchange rate to certain individuals and institutions regardless of types of products, while a residual is subject to strict quantitative restrictions. This system, made possible by the oil revenues, has resulted in heavy reliance on consumer imports to the detriment of domestic production of both importable and exportable goods. 27. Another feature of Angola's trade regime is the existence of "own-funds" imports. One window is provided by the access to extremely cheap air fares, which allows Angolans to travel abroad and purchase imported goods which they resell at parallel market prices. Another own-funds imports window operates through the duty-free shops, where customers may buy imports with dollars obtained outside official channels. Given the distorted trade and pricing regimes and the lack of an appropriate business environment, the own-funds market has a proconsumption bias. The coexistence of liberal windows that channel duty- free consumer imports into the parallel market and strict quantitative restrictions on inputs strongly suggests that effective protection is negative for several industries. 28. Internally, price controls are linked with consumption rationing and wage policy. As noted earlier, the most important component of formal sector wages is paid in the form of buying rights. Although official prices are supposed to apply to all goods and services other than fruits and vegetables, in practice parallel prices prevail and determine consumption patterns in Angola. Particularly because the large oil surplus is allocated at the official exchange rate, the resource misallocation is substantial. At the micro-level, rent- seeking activities divert resources from productive activities and encourage corruption. In addition, the volatility of relative prices in the parallel market discourages potentially productive activities. Nevertheless, parallel markets have helped, particularly in urban areas, to alleviate the distortions created by the official market, have acted as a safety valve to reduce inefficiencies generated by the administrative system, and could be important in facilitating policy adjustment. 29. Consequently, income distribution is characterized by the gap between the urban formal sector and the rest of the country. Workers in the urban formal sector have access to buying rights and therefore to subsidized imports, while the remaining great majority of the population (rural and informal urban) are effectively left outside the system. 30. Public Finance. The public sector comprises the government and about 500 state enterprises. This report focuses on the central government and some major enterprises for which data are available. To obtain a clear fiscal picture, it is necessary to transform revenues and expenditures into parallel market prices. If GDP is also measured at market values, the government deficit for 1989 is estimated at about 12 percent of GDP, compared to 22 percent when measured at official prices. In the case of public enterprises, the consolidated account of the 13 largest shows an operating surplus of 7 percent of GDP at official prices, but just one percent at market prices. Little is known about the other parastatals, but it is likely that they are incurring losses. 31. If the substantial deficit is to be reduced, the wage bill and military expenditures would need to be cut, since together they constitute about 50 percent of total expenditure. 32. Wage Remonetization. The most important components of income in the formal sector are buying rights that provide access to goods at low official prices. Buying rights are therefore instruments of rationing associated with present pricing policies, aimed primarily at protecting the real incomes of  - xv - formal sector workers. Wage remonetization would entail replacement of all buying rights with their money equivalent. As a broad order of magnitude, a budget-neutral remonetization of wages using parallel market prices would require salaries to be multiplied by a factor of 25 for managers and 13 for unskilled workers. 33. It should be made clear that remonetization is a relevant issue only for the 800,000 people in the formal sector; the other 9 million Angolans would remain relatively unaffected. In addition to its other advantages, wage remonetization would reduce the high transaction costs in the present system and make the existing wage structure more transparent, thereby allowing coherent wage policies to be formulated and implemented. Carrying out the remonetization will require improvement of the statistical base on formal' ployment, elaboration of a reliable retail price index, and improvement of payment procedures. The linkage with price liberalization for parastatals will require attention as well. 34. The main analytical findings and recommendations are summarized below. MAIN ANALYTICAL FINDINGS AND POLICY RECOMMENDATIONS Trade and prices Public finance Wage remonetization Changes in official For accurate measure- Buying rights are prices will not nec- ment, public sector instruments of income essarily cause accounts must be re- distribution. changes in parallel formulated using a prices. common numeraire. Wage remonetization would affect only the Price controls should Revaluation of the formal urban sector. be eliminated for all budget using market nonofficial activi- prices leads to a Wage remonetization ties. lower deficit than would reduce transac- the official figure. tion costs, improve Competition should be transparency, and introduced in the Revaluation of the facilitate incomes distribution network. parastatal accounts policy. using market prices Licensing procedures leads to a lower sur- To fully remonetize for new business plus than the off i- present incomes, sal- should be sioplified. cial figures. aries should be mul- tiplied by a factor Privileged access to Exchange rate devalu- of 25 for managers foreign exchange and ation will lead to an and 13 for unskilled the operation of the improvement in the workers. de facto system of recorded government "own-funds" imports finances and a wors- Remonetization should generate a strong ening of parastatal be synchronized with proconsumption bias. sector finances. price liberalization. Effective protection Reduction of the def- As a safety net, some is negative for sev- icit will require buying rights could eral industries. some cut in both the be maintained for wage bill and mili- certain segments of The present foreign tary spending. society. exchange and trade regime is regressive as well as ineffi- cient.  CNATZR 1 Tax TMRDZ flGI3 1.1 Angola's trade regime is characterized by severe distortions that have been costly to the country in several ways. They have had negative distributional effects: access to imported,consumer goods appears to be highly regressive. In addition, they have had perverse microeconomic effects, including incentives to rent-seeking activities and a very distorted price structure. Finally, they have had negative macroeconomic effects since the system encourages imports of consumer goods. 1.2 This chapter describes the salient features of the Angolan trade regime (both external and internal) and derives some policy options for consideration by the Angolan government. 1.3 At first glance, Angola's trade policies appear to be similar to those adopted in other African countries since the early 1970s. A gross currency overvaluation, resulting mainly from inappropriate policy responses to external imbalances, has been accompanied by the imposition of exchange controls. Internally, these policies have been supported by price controls and consumption rationing. As a result, parallel activities have developed in both the foreign exchange and goods markets. 1.4 Angola's trade regime deviates from this standard package in one important way: despite the overvalued exchange rate, several types of liberal foreign exchange windows (free of quantitative restrictions) exist. The foreign exchange allocation system allows flexibility to certain institutions (oil companies, the army) with regard to the type of products they buy, while the residual is subject to strict quantitative restrictions (QRs), which are applicable mostly to inputs. In addition to the inefficient allocation of resources associated with this system, its distributional implications have been largely regressive. 1.5 Given this context, policy changes ought to be guided by four broad objectives: market integration, economic restructuring of the nonoil sector, protection of vulnerable groups, and macroeconomic stability. Price movements, including those of the exchange rate, are the key to market integration since the segmentation of markets is a phenomenon associated principally with the price differential between the official and parallel markets. A more appropriate exchange rate policy is crucial to rebuilding the nonoil sector. Protection of vulnerable groups could be achieved through a remonetization of the wages of certain workers in the formal sector, in a way that secures access to basic items, and through other measures aimed at protecting groups in the informal sector. More study of these issues might be needed. Finally, to attain economic stability, inflationary pressures and sharp real income fluctuations will need to be avoided during the transition period since they could endanger adjustment. 11.6 This chapter has four parts. Seetion I provides a perspective on trade regimes in Africa from the 1970s to the mid-1980s and places Angola in context within a simplified analytical policy framework. Section II discusses the main components of the Angolan trade regime, focusing on the foreign exchange allocation system and the roles of the official and parallel markets. Section III briefly presents the distri-.utional and economic implications of the trade regime. Section IV proposes elements that could be included in a reform of the external and internal trade regimes. Z. A SHPLIX=3D ANLAYSIS F0 !B 5RD RTGIE Z M__ WRC 1.7 During the late 1970s and early 1980s, the exchange rates for many African economies became increasingly overvalued. This overvaluing usually resulted from inadequate macro-policy responses to external shocks (such as  -- 2 - commodity price fluctuations) and to deep-rooted structural problems (such as a weak export base) associated with inflationary budgetary and monetary policies. The reluctance to use the exchange rate as a corrective policy measure eventually led to balance of payments problems and ever greater currency appreciation in real terms. By the mid-1980s, quite large parallel markets had developed in many countries in which foreign exchange was being traded at a significant premium over the official rate. 1,8 Once passive exchange rate management was taken as given, the policy response to the overvalued official currency was internally consistent in most countries, at least in principle. The overvalued currency generated excess demand for imports at the official rate, which required the introduction of QRs. QRs, in turn, implied that the market could no longer be relied on to allocate resources, and an administrative system of foreign exchange allocation evolved. The overvalued currency and the subsequent administrative allocation of foreign exchange (QRs) meant that potentially large rents and profits were available to those with access to cheap foreign exchange. Internally, the overvalued currency and foreign exchange rationing had to be supported by price controls and some kind of consumption rationing. By the late 1970s, most of these policies had become prevalent in trade regimes in Africa. 1.9 Within their well-known limitations (antiexport bias, misallocation of imports), the "effectiveness" of these restrictive trade regimes varied widely, with correspondingly divergent economic outcomes in different countries. Countries like Kenya, Malawi, and Zimbabwe managed, despite their restrictive trade regimes, to avoid major and sustained misallocation of resources, and so to avoid a decline in economic activity (Table 1.1). Most probably, this was due to their ability to maintain the exchange rate overvaluation within moderate limits and to their relatively strong institutional and civil service capabilities. These factors, especially the lower currency overvaluation, translated into better macroeconomic management and less domestic and external instability than in other countries. 1.10 At the other extreme, countries like Ghana, Mozambique, and Tanzania experienced unprecedented economic declines from the late 1970s to the mid-1980s, due largely to the severe resource misallocations caused by the exchange rate and trade regimes. A large overvaluation of the currency implied severe distortion of the overall system of prices and incentives, while weak institutional capabilities rendered the centralized allocation of resources totally ineffective. In Tanzania, for example, a firm-by-firm survey revealed that the centralized administrative allocation of foreign exchange was worse than random: inefficient enterprises benefited disproportionately. 1.11 By the mid-1980s, declining GDPs, balance of payments crises, and high inflation forced some of the worst performers to embark on radical reform programs. Ghana and Tanzania, among others, dropped many QRs and price controls and implemented policies that realigned their exchange rate, thereby allowing market forces to play a key role in allocating resources. Tanzania embarked on a major program of internal trade liberalization as well. The emerging evidence indicates that these reform programs are significantly boosting GDP growth and consumption per capita. 1.12 A more realistic exchange rate regime and internal and external trade liberalization measures have had a number of implications for foreign exchange allocation. In Tanzania, for instance, the gradual movement of the official exchange rate toward more realistic levels has permitted the implementation of a gradually expanding open generalized licensing (OGL) facility. The OGL allows any importer automatic access to foreign exchange (at the official rate) to import any good that has been included on the (expanding) OGL list. Furthermore, since 1984, the government has been implementing a policy of "own-funds" imports that allows importers to obtain a license to import virtually any good provided they use their own foreign exchange - and no questions are asked regarding the source of these funds. The positive impact of this measure alone has been enormous: imports rose rapidly and so did the availability of critical goods in  -3- TABLE 1.1 KEY ECONOMIC INDICATORS IN SELECTED SUB-SAHARAN AFRICAN COUNTRIES. 1982 Annual GDP REERa/ Budget CPI growth Country (1972=100) deficit/GDP (1972=100) (1972-82) Group I Kenya 102.40 -7.0 379 6.3 Malawi 82.36 -8.0 262 4.1 Zimbabwe 86.00 -11.0 243 4.3 Group II Ghana 1,252.21 -7.0 730 -2.0 Mozambique n.a. -10.1 n.a. -9. oi/ Tanzania 166.86 -9.0 327 1.3 Angola n.a. -22.0 n.a. -0.8 a. Real effective exchange rate: above 100 implies overvaluation of the local currency. b. Refers to 1980-86. Source: Exchange Rates and Parallel Market Economies in Sub-Saharan Africa, World Bank Staff Paper No 11, IFS; data for Angola are Mission estimates. rural areas, resulting in a substantial supply response and a tapering of f of inflationary pressures. 1.13 Mozambique, which has experienced a long period of war similar to Angola' s and also attempted to introduce a centrally planned economy, implemented a major adjustment program somewhat later than Tanzania. Economic growth has resumed after a long period of economic decline. The main elements of the program, launched in 1987, include substantial exchange rate and price adjustments, which led to a drop in the nominal parallel market exchange rate from 1,800 meticals per U.S. dollar in 1987 to 1,250 in late 1988. Unification of the exchange rate market has not yet been completed, and by end-1989 the parallel rate (at 1,950) was still double the official rate. 11. MM ANGOLA TRDI RGZ_E= 1.14 Angola's trade regime seems to fit within the policy package adopted by many African countries since the early 1970s. And given the magnitude of its currency overvaluation and its institutional and economic management weaknesses, Angola might be said to fall in the category of countries like Ghana, Nozambique, or Tanzania in the early 19809. 1.15 The degree of currency overvaluation in Angola is suggested by the enormous gap between the official and the parallel exchange rate. The official exchange rate has remained fixed since Independence in 1975, and in April 1991 it was 1/100th the parallel rate. By comparison, in Mozambique, the official rate was about 1/50th the parallel market rate just before implementation of the exchange rate reform, while in Ghana it was 1/20th the parallel rate at the peak of the economic crisis. 1.16 The underlying factor behind the overvaluation of the kwanza has been the monetization of large budget deficits, recently estimated at around 20-25  - 4 - percent of GDP.1" The Angolan authorities, however, inspired by the principles of central planning, resorted to administrative controls to contain the pressures arising from the expansion in domestic demand and the declining trend in the supply of nonoil products. Thus from the outset, the policy response has been direct intervention in the allocation of resources, including, among others, foreign exchange rationing (import QRs), price controls on most goods and services, consumption rationing through the distribution of buying rights that primarily benefit formal sector workers in the cities, and centralized control of internal distribution. 1.17 The acute distortions generated by the overvalued exchange rate and by the administrative system introduced to deal with the macroeconomic disequilibrium fueled the growth of an increasingly important parallel economy. Aside from some commercial agriculture, the parallel market for goods has concentrated on trading activities in the cities. An incipient service sector began to develop during 1990-91. 1.18 This all too familiar picture is, however, transformed by the existence of virtual OGL windows (imports with no QRs) in Angola despite the vastly overvalued exchange rate. As explained below, the foreign exchange allocation system in Angola allows for varying degreei of flexibility in applying licensing requirements and QRs to different socioeconomic actors and hence to different categories of goods. This practice constitutes an important deviation from the standard policy depicted above and critically influences the functioning of the Angolan economy. A. External Trade 1. Foreion Exchanae Allocation System 1.19 At about US$300 per capita, Angola's merchandise exports compare favorably with those of other middle-income oil-exporting African countries like Nigeria (US$157) and Cameroon (US$146), and are the same as those of Argentina. Despite these relatively high export earnings, Angola was faced with an increasingly tight foreign exchange constraint throughout the 1980s. 1.20 Under Angola's foreign exchange allocation system, virtual OGL windows coexist with QRs of varying degrees, a contrast with the very restrictive QRs adopted in other countries such as Ghana (1972-83), Kenya (1973-80), and Tanzania (1967-83). The pattern of QRs in Angola, when effectively in place, has also been fairly atypical. In most countries, QRs have taken the form of prohibitions and ceilings on imports competing with local industry and on imports of nonessentials. In Angola, QRs have generally had the opposite effect of discouraging local production. 1.21 How can virtual OGL windows be sustained when the official exchange rate is only 1/100th of the parallel rate, however nontransparent and camouflaged the OGL windows? First, in practice, there is a variable limait to demand through the liberal import windows (OGLs, limited general licenses; see Table 1.2 for definitions and details), which is determined by some aggregate (and flexible) ceiling and by the small size of the eligible population (the urban upper class and some of the middle class). Second, the sustainability of the high- consumption import windows has been assured through the very large oil receipts. If, say, 150,000 Luanda urbanites have access to OGL consumption goods, which would use, say, US$300 million a year (only about 15 percent of total oilrevenue), that would allow each of them US$2, 000 worth of goods a year. This is a very perverse case of "Dutch disease." This perverse and internally inconsistent trade policy regime may have evolved from a more conventional Dutch disease syndrome that was made worse by economic management held hostage to the 1/ At official prices.  TABLE 1.2 MLTRIX FOR ANALYSIS OF THE UNGXLAN FOEMIGN EXCHANGE ALLCATION SYSTEM Foreign effective exchange exchange Degree comments/ window ~Main market Main actor -rate Type of goods of QgR/ Isue OGLi? oil * Company shops * Oil firma * Parallel * Oil machinery Small Official ex- company for consumer * Elite/formal * Official * Consumer change rate ap- goods urban workers goods, includ- plies to con- *Direct use of ing luxury sumer goods, inputs parallel rate applies to in- puts purchased from retained exports OGL army * Direct sloca- * Army * Official * Machinery Small The share of tion of capital * Soldiers * Consumer goods the army allo- and inputs cation for con- * Army shops sumer goods is unknown LWL" (a) * Duty-free shops * Priority para- * Official * Luxury con- Small LGL may be a * TAAG trips statals (SONAN- sumer goods virtual OGL if GAL, ENDIAMA) overall ceil- ings are non- binding LGL (b) * Lojae * Mid-level * Official * Nonbasic Moderate Cartao com- complementarlds civil servants consumer goods plementarlo, LGL (b) some- what more re- strictive than for LGL (a) MGLdU * Lojas * Urban formal * Official * Basic goods Large Cartao baslco, abasteclmlento workers quantitative rationing Aulp/ * Inputs, direct * Firms * industrial Very What is actual allocation to * Rural shops inputs large basis for al- industries * Others location of inputs? OF91 urban * Parallel markets * Informal sector * Parallel * Consumer goods Small Possible main Duty-free shops * Urban poor sources of own funds: unoffi- cial exports, repatriated capital a. In this column, an attempt is made to suggest the degree of "restrictiveness" of quantitative restrictions (OR) for particular purposes (or foreign exchange windows). "Very large" implies a high degree of restrictiveness, while "small" implies that the system functions like a virtual OGL system. b. OGL, or open generalized licensing: Aggregate limaits are flexible and virtually nonbinding. Similarly, product- or beneficiary-specific limits on effective foreign exchange licensing are nonbinding or nonexistent. c. LGL, or limited generalized licensing: Aggregate limits may be binding while product-specific limits do not effectively exist. For some, beneficiary-specific limits apply (those who have access to cartaos complemnentarlaa, foreign exchange to travel with TAAG, and the like). d. GAL, or generalized administrative licensingt Aggregate limits are more binding than in LGL, and product-specific limits are effective (cartaos basicos). e. AL, or administrative licensing: Aggregate and product- or beneficiary-specific limits apply in effect (largely the result of its residual nature). f. OF, or own-funds scheme: Imports financed out of foreign exchange purchased in the parallel market or acquired through unofficial exports of capital repatriation. war eff?rt, urban needs, and the relatively recent Portuguese colonial legacy. 1.22 Anotier unusual feature of Angola's trade regime is the "own-funds" import windows. They do not appear to be the result of a conscious policy decision, as they were in Tanzania, but rather seem to have evolved as a creative 2J Incidentally, in the pre-Independence days, Portugal used to dump cheap liquor in Angola to support its own industry.  - 6 - response by economic agents to the immense distortions and loopholes in the kngolan economy. One such window operates through the Angolan airline TAAG, which charges US$15 round-trip (at the parallel exchange rate) to Portugal or Brazil, where passengers can buy large quantities of imports.3/ Another operates through duty-tree shops, which allow customers to buy imported goods upon presentation of foreign exchange alone, presumably purchased in the parallelmarket or acquired through illegal exports or capital repatriation. The own-funds imports schemes are not institutionalized, and they operate through a few narrow channels. Given the distorted internal trade and pricing regimes and the skewed purchasing power in Angola, the own-funds market also has a pronounced proconsumption bias. 1.23 To understand the Angolan trade regime and associated policies, it is helpful to identify the chief economic actors in Angola: (1) the foreign (oil) firm, (2) the military, (3) the urban elite (employees of priority parastatals, high-level technocrats), (4) the urban formal sector worker (civil servants, other firm workers), (5) the urban poor, (6) the farmer, and (7) the domestic firm. Each window of foreign exchange allocation may be linked to particular actors and to other important characteristics such as the degree of restrictiveness of the QR regime for each window or the effective exchange rate. 1.24 Official Foreign Exchanae Allocation. The bulk of official foreign exchange comes from the oil receipts generated in the Angolan offshore fields by foreign enclaves. Prom the US$2.3 billion of oil exported every year, 4/ the foreign enclaves retain almost one-third to cover external payments, profit remittances, and consumer goods to stock the companies' shops. Therefore, for all practical purposes, these oil companies operate under an OGL system that allows unlimited imports of any type of product, including consumer goods that are sold at the official exchange rate. 1.25 In principle, the Banco Nacional of Angola (BNA), the Angolan central bank, allocates the remaining oil receipts according to a foreign exchange budget approved by the Ministry of Planning (see Introductory Economic Review, UNDP/World Bank, June 1989). The degree of restrictiveness and accessibility varies according to the recipient or institution. 1.26 Imports for the army effectively fall into an OGL regime since aggregate limits are flexible and virtually r.ouinding. Foreign exchange outlays by the military, including imports of consumer goods, are estimated at over US$500 million (excluding those financed by official transfers), which represents over a quarter of total oil revenues. The share of consumer imports for the military is estimated at 20 percent (or US$100 million). 1.27 Certain priority companies, such as Sonangol (the oil parastatal), Endiama (diamonds parastatal), TAAG (Angolan airline), duty-free shops, and the Hotel Presidente, also receive special treatment. Overall ceilings on this use of foreign exchange are set at the beginning of each year, but they are quite flexible, and access to foreign excharge is more or less automatic. Since these companies can bring in any goods they want, they are essentially subject to limited generalized licensing (LGL). These imports are not recorded in the balance of payments as merchandise i Sorts, so no estimate of the quantities imported in this way is possible.5/ 3 According to recent requlations, Angclan nationals are entitled to US$500-1,000 travel allowance within a four-year period. The beneficiaries are about 600,000 employees, some of whom can use the allowance either to travel or to purchase goods inside Angola. Thus, they could bring in imports totaling US$75-150 million a year, at the official rate. 4/ Official exports of diamonds amount to over US$200 million; other exports (coffee, for example) are negligible. 5 This problem will be corrected when the new balance of payments methodology pr^posed by the IMF is adopted.  1.28 In practice, the amount of foreign exchange available for imports under direct administrative allocation is a residual. In 1989, it was an estimated US$700 million, excluding capital goods imports.61 Even this narrowly defined level of civil imports is higher than total imports for many other Sub-Saharan countries in per capita terms (Table 1.3). 1.29 Within this framework of administrative allocation, there are also different degrees of flexibility according to the type of products or recipients. Priority is given to import licensing for consumer goods to meet urban demand, although the QR is stricter than that faced by the privileged institutions described above. In particular, imports of items to cover the general quantitative basic entitlement for formal sector workers are subject to stricter QRs than those to cover the complementary entitlement (set in value) for technicians and mid-level civil servants. In fact, the allocation of foreign exchange for intermediate goods constitutes the real residual, actually subject to an administrative license in the proper sense (see Table 1.2). With the partial data available, a preliminary picture can be constructed of how the government allocates foreign exchange among types of products and sectors (Table 1.4). 2. Structure of Imports 1.30 Consumer Goods. Consumer goods as a share of noncapital merchandise imports rose from 27 percent in 1985 to 45 percent in 1988-89. Of the US$287 million allocated to imports of consumer goods in 1988, an estimated US$200 million was for basic food imports. (The amount of basic food imports corresponds roughly to the value of the basic and complementary entitlements - cartaos - of formal sector workers mentioned above.) 1.31 In 1988, food imports accounted for around 15 percent of total merchandise imports of US$1,399 million. This share is too high for a country endowed with large agricultural resources and in which agriculture accounts for about 15 percent of GDP. In Zambia, by contrast, food imports account for only 4 percent of total merchandise imports, while the share of agriculture in GDP is only 11 percent. Furthermore, these food imports in Angola serve mainly the urban population, who account for about 30-40 percent of the population. 1.32 Intermediate Goods. The share of official foreign exchange allocated to imports of intermediate goods has been declining in relative terms, dropping from 58 percent in 1985 to 40 percent in 1989. Intermediate goods for agriculture accounted for barely 3 percent over the period 1985-89, and their share fell to 2 percent in the 1990 foreign exchange budget (Table 1.4). 1.33 The share allocated to manufacturing inputs has also declined, falling from 23 percent in 1985 to 18 percent in 1989. Foreign exchange seems to be allocated to this sector to prevent closures and labor layoffs rather than on the basis of economic efficiency in different industries. Representatives of 18 industrial firms interviewed in November 1989 claimed that capacity utilization is constrained by their limited access to foreign exchange to purchase raw materials and that some factories have had to be temporarily shut down.7/ Furthermore, despite the decline in production over the last decade, the number of workers has not fallen - and in some cases it has even increased. As a result, even with subsidized inputs, unitary production costs are expected to be higher than those of imported goods (at the official exchange rate). 1.34 A disaggregated analysis would be required to establish whether the net value added of some activities would be negative if prices and inputs were valued 6/ Capital imports, estimated at US$252 million in 1988, are apparently financed mainly by foreign investment or long-term borrowing. 71 Capacity utilization ranged from 10 percent to just below 50 percent. See informal paper produced by an industrial sector reconnaissance Mission in December 1989.  TABLE 1.3 PER_ CAPSTA GPAND IORTS la SELECTED (U.8S. dollars) GDP Country 1987 Lmprts/ Ghana 374 72 Kenya 315 85 Malawi 141 35 Mozambique 103 51 Tanzania 129 51 Zimbabwe 582 123 Angolab/ 700 78-/ a. Officially recorded merchandise imports in 1988. b. Preliminary estimate. C. Imports under direct allocation in 1989. Source: World Development Report 1989, World Bank, 1989. at border prices. In any case, it is clear that resource misallocation is widespread. A striking example is that imports of raw materials for agriculture amounted to US$22 million in 1988 while imports of beverages (including beer and whiskey) reached about US$25 million.8/ While the market is flooded with imported beer, some local breweries have closed because of a lack of raw material imports. wheat flour factories have encountered a similar problem because of a shortage of imported wheat. 1.35 Imalications. The virtual OGL/LGL windows in Angola benefit from the extremely overvalued official exchange rate. Thus, the theoretical internal consistency of the administrative regime is not being achieved in practice in Angola, and the ensuing distributional consequences and the misallocation of resources are made even worse by the nature of these OGL facilities. Several implications follow: * The system is biased toward consumer goods, especially luxury items (beer, spirits), through the operation of both the OGBJLGL windows and the more restrictive administrative licensing system. * There is a bias against intermediate goods for local production, which is reflected in a disproportionately low overall allocation to agriculture compared with the sector's output potential. In addition, although the overall level of foreign exchange allocated to manufacturing may be appropriate, it appears to be distributed thinly across firms in order to enable them to survive at low capacity, without reference to potential competitiveness. / This figure underestimates beverage imports because it is based on the partial records of the Societe Generale de Surveillanoe (SGS), which exclude imports for the army. 9/ This is in sharp contrast to the more conventional and internally consistent restrictive trade regimes, where the OR and administrative allocation systems usually ensure that the importation of nonbasic consumer goods is minimized.  -9- TAWS6 1.4 fg ;QII-MG A T12 (Mollions Of U.S. dolars) 1990 Imprt 1985 1986 1987 1988 19098/ Budget Consumer goods 157.7 74.7 134.4 287.0 236.1 287.1 Intermediate goods 333.7 138.1 155.7 269.4 211.1 356.8 Agriculture 15.9 9.9 13.4 21.7 16.9 18.0 Manufacturing 134.6 37.0 60.7 121.4 96.1 167.0 Petroleum 77.9 57.5 39.5 59.3 27.0 60.0 Snergy 7.6 2.1 5.0 7.6 5.1 8.0 Transportation and 41.9 15.9 14.8 21.5 30.0 25.0 communication Social consumptionb/ 87.5 24.8 45.1 78.9 74.1 126.1 Total 579.0 237.6 335.1 635.3 521.5 770.0 a. To September. b. Includes medical and educational supplies. Source: Banco Nacional de Angola (SNA). - There is also a bias against domestic production of many products, as illustrated by the beer and wheat flour industries. Unlike the experience in other countries, import restrictions do not appear to be directed at protecting local production, whether for import substitution or for export. Although the official overvaluation of the currency underlies the basic disequilibrium in the foreign exchange market, the foreign exchange allocation system has contributed to this disequilibrium through its adverse impact on domestic production. * Although the size of the foreign exchange allocation to the army is not an economic issue, by and large, it may well be that the prevailing trade and price regimes encourage the purchase of expensive equipment and nonessential consumer goods, while imports of intermediate goods suffer. I