Report No. 7795SMOZ Mozambique Industrial Sector Study The Development of Industrial Policy and Reform of the Business Environment May 22, 1990 Southeem Africa Department Industry and Energy Operations Division FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in dhe performance of their official duties. Its contents may not odterwise be disclosed widout World Bank authorization. 'J CURRENCY EQUIVALENTS (AFRIL 1990) US$1 841 Neticais (MT) - April 1990 MT 1 - US$ 0.001 ABBREVIATIONS AND ACRONYMS AEPRIMO Association of Private Sector Entrepreneurs DM Banco de Mocambique BPD Banco Popular de Desenvolvemento BST Banco Standard Totta BSTH Banco Standard Totta de Mocambique CCADR Caixa de Credito Agrario e Desenvolvimento de Seguros CNP Conmissao National de Planc DRC Domestic Resource Cost EGA Custom Fees EPR Effective Protection Rate ERP Economic Rehabilitation Program GAPI Cabinete de Consultoria dos projectos da Pequena Industria GPIE Foreign Investment Promotion Bureau IDIL Instituto de Desenvolvimento da Industria Local MIE Ministry of Industry and Energy OER Official Exchange Rate SEILA Secretariat of State for light Industry SER Shadow Exchange Rate SME Small Medium Enterprise SNAAD System for Non Administered Allocation of Foreign Exchange SOCIEF Sociedade de Investimentos e Estudos Financeiros UCPI Coordination Unit for Industry Import Programs Fiscal Year Government and Publit Fiterprises: Calendar Year FOR OFFICIUL USE ONLY MOZANBIqUE INDUSTRIL SECTOR STUDY PREFACE This report is based on a study of the business environment carried out by the IDA in conjunction with UNIDO and financed by UNDP. The study was submitted to the Government in July 1989, and discussions of its findings were held in Maputo in December 1989. A number of changes were made as a result of these discussions. as well as updating and ext.nsion. The initial study team consisted of Messrs. David Phillips (IDA and principal author), J. Weiss, D. Rhatigan, and H. Jackelen (Consultants). Contributors to the study in IDA included Mr. C. Frischtak (IENIN), who was lead advisor, and valuable assistance was received from AF6CO, AFTTF, and colleagues in AF6IE. The main mission spent four weeks in the field over the period November to December 1988, followed by a further visit in March 1989. In December 1989, during the discussions with the Government, the study was updated in preparation for this final report. The study serves as the policy basis for two current IDA investment projects, the Small and Medium Enterprise Development, and Industrial Enterprise Restructuring Projects, which are due to become effective in mid- 1990. It also complements the UNIDO Industrial Policy and Institutional Adjustment Project, which is notr under way. The discussion is concerned with the development of industry sector-wide, but, in accordance with its original terms of refereuice, it particularly focuses on small scale industry. The mission would like to express its appreciation for the help and guidance received from the Government, in particular H.E. Antonio Branco. Minister of Industry, Mr. 0. Mutemba. Vice Minister, and his staff, and staff of the Bank of Mozambique and the Ministries of Commerce and Finance, who participated in the study discussions upon which the report is based. We would also like to express our appreciation to Mr. Charles Larsimont, Resident Representative, UNDP, and Mr. Carlos Goulart of UNIDO, and many individuals in both the private and public sector and aid agencies, who gave up their time to assist the mission. This document has a restricted distribution and may be used by recipients only in the performance | of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. NDZAMBIQUE INDUSTRIL SECTOR STODY TIE DEELOPMENT 0? INDUSTRIAL POLICY AND REFORM OF THE BUSIESS EIR01ONMENT TABLE OF COrTENTS Page No. PAI. BACKGROND, SUMRY AND RECOIHENDATIONS . . . . . . . . . . 1 A. Background and Objectives . . . . . . . . . . . . . . 1 B. Conclusions, and Reconmendations . . . . . . . . . . . 5 -Industrial Structure . . . . . . . . . . . . . . . I 5 -Economic Efficiency of Industries . . . . . . . . . . 6 -Industrial Sector Po.Acy . . . . . . . . . . . . 6 -Financial Sector Policy . . . . . . . . . . . . . . . 8 i II. ECONOMIC BACKGROUND AND INDUSTRIAL GROWN . . . . . . . . . 10 A. General Economic Background . . . . .. . . . . . .. 10 B. The Development of the Industrial Sector . . . . . . . 11 - ludustrial Development: Preindependence . . . . . . 11 - Industrial Declines Post-independence . . . . . . . 13 - Manufacturing Employment . . . . . . . . . . . . . . 16 - Export of Processed Goods . . . . . . . . . . . . . 17 C. The Structure of Industry . . . . . . . . . . . . . . 17 - The Composition of Industrial Output . . . . . . . . 17 - Industrial Location . .19 - Ownership Structure of Industry . . . . . . . . . . 20 D. Industrial Strategy and Planning 1977 - 87 . . . . . . 20 E. Small Scale Enterprises; Structure and Performance . . 21 III. TE CURRENT BSINESS ENVIRONHENT . . ...... . * . . 25 A. The Business and Investment Climate . . . . . . . 25 - Generl Background . .25 - New Investment and Company Formation . . . . . . . . 26 - Enterprise Pricing..... . . .28 - Marketing and Distribution . . . . . . . . . . . . 29 - Capacity Utilization . . . . . . . . . . . . . . . 30 - Current Industrial Output and Employment . . . . . . 30 - Manufactured Exports . . . . . . . e . . . . . . . . 32 B. The Demand for Industrial Products . . . . . . . . . . 32 - Long Term Demand . . . . ... . . . . . . 33 - Impact of Structural Adjustment on Purchasing Power 34 - Impact of Donor Aid on Production and Markets . . . 38 - ii - C. The Reform of Industrial Policy . . . . . . . . . . . 38 - Reforms Under the Economic Rehabilitation Program . 38 - Reforms and Medium Term Strategy for Industry . . . 42 IT. ESTIMATION OF THE ECONOMIC EPPiCIENCY OF INDUSTRY . . . . . 64 A. Industrial Efficiency Analysis . . . . . . . . . . . . 44 S. The Enterprise Survey, Industrial Efficiency Estimates 46 C. Results of the Survey; Industrial Efficiency Estimates 46 D. Conclusions ..... ... ..... , .49 V. TEE DEVELOPMENT OF INDUSTRUL POLICY . . . . . . . . . . . 54 A. Introduction . . . . . . . . .. . . . . . . . . . . 54 B. Exchange Rate Policy . . . . 55 _ Objectives of Exchange Rate Reform . . . . . . . . . 55 - Effects of Currency Devaluation on Efficiency and Exports . . . . . . . . . . . . . . .57 - Equilibrium Exchange Ratet . . . . . . . . . 58 C. Foreign Exchange Allocation . . . . . . . . . . . . . 59 - Channels of Foreign Exchange Provision . . . . . . . 59 - The Export Retention Scheme . . . . . . . . . . . . 60 - Donor Import Allocation Procedure . . . . . . . . . 61 - Guidelines for Donor Import Allocation System . . . 63 The Small-medium Erterprise and Market Fund . . . . 65 - Import Licenses . . . . .65 Revised Guidelines for Import Licenses . . . . . . . 66 - S-umary of Recommendations on Foreign Exchange Allocation . . . . . . . . . . . 67 D. Trade Tariff Reform 68 - The Current Reform Program. .. . 68 _ Tax Revenue Considerations . . . . . . . . . . . . . 69 - Designing the Tariff Structure . . . . . . . . . . . 70 - Income Distribution and Tariff Policy: Luxury Goods 72 - Tariffs and Long-run Protection . . . . . . . . . . 73 - Export Promotion and Tariff Policy . . . . . . . . . 75 - Summary of Tariff Reform Recommendations . . . . . . 76 E. Measures for Price Setting . . . . . . . . . . . . . . 78 - Conditional Prices . 78 _ Revision of Conditional Price Formula . . . . . . . 79 - Demand-based Pr cing. 81 - Summary of Recommendation on Conditional Pricing . . 81 P. Investment Incentives 81 - The Current System. . . ..81 _ iii - - Rationalization of Tax Exemptions . . . . . . . . . 82 - Employment Tax Credit . . . . . . . . . . . . . . . 84 - The Effects of the Turnover Tax (Imposto Ctrculacao) 85 - Access by Small Enterprises to Tax Incentives . . . 86 - Effects of P.oposed Tax Changes . . . . . . . . . . 86 - Summary of Recommendations on Taxation . . . . . . . 87 VI. THE WIKT.NG SECTOR AND TEE FIZWICING OF INDUSTRIAL . . . . 88 REHABILITATION A. The Role of the Financial Sector . . . . . . . . . . . 88 B. The Operation of Banking Institutions in Mozambique . 89 C. The Reform of the Banking Sector . . . . . . . . . . . 92 - Economic Stabilization: Credit Ceilings . . . . . . 92 - Structural Reforms in the Financial. System . . . . . 93 D. The Demand for Industrial Credit . . . . . . . . . . . 95 - Real Domestic Lending . . . . . . . . . . . . . . . 95 - Industrial Credit Survey . . . . . . . . . . . . . . 96 E. Enterprise Restructuring; Revaluation of Assets . . . 98 - objectives . . . . . . . . . . . . . . . . . . . . . 98 - Asset Revaluation: A Case Study . . . . . . . . . . 99 VII. TIE PROKMOTION OF INDUSTRIAL INVESTHENT . . . . . . 102 A. The Environment for Foreign Investment . . . . . . . . 102 B. Foreign Exchange Allocation; Export Retention Scheme 103 C. Debt Equity Swap Law; Effect on Investment Incentives 105 D. Industrial Promotion Infrastructure . . . . . . . . . 108 - The Role of the Unidades de DiregiLo . . . . . . . . 108 - The Promotion of Small Scale Enterprises . . . . . . 110 E. Development and Diversification of Industrial Financing 112 - The Environment for Equ ty Investment . . . . . . . 112 - Developments in Financial Service and Markets . . . 113 - Resources of Entrepreneurship in Mczambique . . . . 114 - The Industrial Labor Force: Share Participation Schemes . . . . . . . . . . . . . . . . . . . . . . 115 - The Role of Venture Capital . . . . . . 116 - Management Legal and Accounting Services .118 F. Mechanisms for Privatization/Transfer of Companies . . 119 G. Institutional and Management Considerations for Equity Financing ....... 121 H. Conclusions . . . . . . . . . . . . . . . . . . . . . 124 - iv _ TABES. * * e a * * a a * * a a a * a a *. * . v 2.1 Preindependence Growth of Enterprises and Employment . . . 13 2.2 Output Volumes: Power and Principal Manufactures . . . . . 15 2.3 Growth of Manufacturing Employment . . . a . . . . . . . 16 2.4 Processed Export Products 1973-87 . . . . . . . . a . . . . 17 2.5 Structure of Manufacturing Output . . . . . . . . . . . . . 17 2.6 Level of Manufacturing Value-Added (1973) Selected Low Income African Economies aa... 18 2.7 Ownership Structure of National Enterprises . . . a a 20 3.1 Foreign Investment Proposals Approved by GPIE . . . . . . . 26 3.2 Foreign Investment Approvals by Subsector . . . . . . . . . 26 3.3 New Enterprise Registrations in Small-Scale Manufacturing Processing and Repairs Maputo . . . . . . . . . . . . . . . 27 3.4 Capacity Utilization of Selected Companies a. . . . . . . 30 3.5 Output Growth - MIE Enterprises 1988-89 . . . . . . . . . . 31 3.6 Population and Consumption .. . . . . . . .. .. . . . 33 3.7 Consumer Price and Wage Index, End Year . . . . . . . . . . 35 3.8 Enterprises Reporting Demand Constraints . . . . . a . . . 37 3.9 Finished Goods Stocks; Selected Industries . . . . . 37 4.1 Domestic Resource Cost Ratios (Medium Term) . . . . a a a 48 4.2 Domestic Resource Cost Analysis (Short Runj Category A 45 Enterprise Survey . . a .. a ..... .. C.... 51 4.3 Dow3stic Resource Cost Analysis (Short Run) Category B 45 Enterprise Survey .*. . . a. . . .. ...... . .....a 52 4.4 Domestic Resource Cost Analysis (Short Run) 10 Enterprise Survey . . .a .. . a... ... .a.. .. 53 5.1 Foreign Exchange Allocation and Production MIE Enterprises 1989 62 5.2 Iliustrative Grouping of Branches by DRC Results . . . . 65 5.3 Trade Taxes in Total Tax Revenue: Mozambique and Other African Economies 1980s . . . . . . . . . . . . . . . . . . 70 5.4 Tariff Protection and Other Taxes and Subsidies . . . . . . 77 5.5 Increases in Conditional Industrial Prices 1986-88 . . . . 78 5.6 Comparative Output Price Ratios (Late 1988) . . . . . . . 80 5.7 Illustration of Effect of lO Turnover Tax . . . . a. . . 85 6.1 Banking System; Domestic Credit Ceilings and Utilization 92 6.2 Net Domestic Lending to the Economy . . . . . . . . . . . . 95 6.3 Survey Results - Summary Demand for Industrial Credit 1988 97 6.4 Asset Revaluation: Mabor de Mozambique . . . . . . . . . 99 6.5 Coefficients for Revaluation of Fixed Assets . . . . . . 100 FIGURES . * . * . . . . . . . . . . . . . . . . . . . 2.1 Index of Industrial Production . . . . a . . . . . . . . 14 2.2 Industrial Enterprise Employment: Economy-Wide 1973 . . . . 22 2.3 Industrial Enterprise Employment: Economy-Wide 1987 . . . . 23 2.4 Industrial Enterprise Employment: Maputo 1973 . . . . . . . 23 2.5 Industrial Enterprise Employment: Maputo 1987 . . . . . . . 24 5.1 Evolution of Exchange Rate and Prices . . . . . . . a. . . 56 6.1 Shares in New Lending by Banks . . . a . a a . . . . . . 91 1XVIL APEDICES Appendix 4.01. Domestic Resource Cost Measure of Economic Efficiency .. ...................... . 125 Appendix 4.02. Analysis of Efficiency of Industrial Subsectors . 134 Appendix 5.01. Calculation of Effective Protection Rates . . . . 145 Appendix 5.02 Effects of Changes In Tax Incentives . . . . . . . 148 Appendix 6.01 Operating Mechanism for Revised Export Retention Scheme * ........................... . . 151 Appendix 6.02. Company Privatizations Illustration of Porm of Management Contract and Company Transfer Process . . . . 152 AS= . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Summary Accounts of the Banco de Desenvolvimento and the Banco Standard Totta de Mozambique, 1984-89 . . . . . . . . . . . 154 2 Summary Accounts of the Bank of Mozambique, 1984-89 . . 155 3 Structure of Interest Rates, September 1989, Ministerial Orde. ...156 4 Distribution of Bank Credit to the Economy. . . . . . . . . 159 5 Legal Status of Enterprises . . .160 6 Maln Problems Affecting the Firms . . . . . . . . . . . . . 161 7 Import Tariff Rates - 1989 . . .162 8 Export Retention Rates . . . . . . . . . . . . . . . . . . 166 9 Fixed Consumer Prices 1980 thru April 1988 . . . . . 167 10 Produc-ion Data Ministry of Industry and Energy . . . . . . 168 11 Scale Distribution of Industry 1973 and 1987 . . .... . 169 mOzWMBIQUR INDUSTRIAL SECTOR STUDY THE DEVELOPMENT OF INDUSTRIAL POLICY AND REFORM OF THE BUSINESS ENVIRONMENT I. BACKGROUND. SUMMIARY AND RECOMMENDATIONS A. Background and Chjiectives of the Study 1.01 This report was conceived as a study of broad industrial policy within the context of the rapid chantS,. in the business environment brought about by the Economic Rehabilitation rr.gram, launched in January 1987. Its purpose is to identify important medium term options in a range of specific industrial trade and financial policy areas, rather than to carry out a comprehensive analysis of industrial activity and development per se, which has been the subject of other studies. 1/ 1.02 A series of recommendations and proposals are set out which are intended to assist the Government to initiate action and prioritize issues for more detailed study. Revisions and updating have been made following discussions held with the Government on the draft report. The current rapid momentum of policy change in Mozambique has already encompassed some of the recommendations made, while others are under consideration as part of the Government's own ongoing industrial policy formulation process. 1.03 The main co.lclusions are summarized as follows: Industrial Development: - for historical reasons the current pattern of industrial production is consistent with econcmic efficiency and future investment should maintain this broad structure. - existing industrial capacity, If fully rehabilitated from its currently poor condition, may exceed domestic demand in the medium-term. Thus rehabilitation must be selective, not comprehensive. - along with selective rehabilitation of major enterprises, new investment may be justified in expanding micro-enterprises which have shown apparent dynamism in adverse conditions. Industrial and Trade Policy: - foreign exchange allocation and import licensing should, during transition to more opera markets, be based on specific 1, E.g. UNIDO Industrial situation in Mozambique, Vols 1 and 2, May 1987; and Review of Agricultural and Industrial Enterprises, Vol 1 and 2, World Bank (Arthur D Little inc), July 1988. 'Mozambique; an Introductory Economic Survey' World Bank, June 6, 1985, provides an account of the industrialization process before independence. - 2 - international competitiven%3s and price criteria, phasing out administrative control. - the import tariff and control system should continue to be reformed, using concepts of transparency and uniformity. monitored by a tariff commission. - the corporate tax system could be rationalized to improve incentives for existing businesses, while at the same time maintaining the level of Government revenue. Finance and Investment Promotions - the impact of investment promotion instruments must be continuously monitored to ensure effectiveness; the industrial promotion institutions should aim to move towards financial self-sufficiency through increased autonomy and provision of fee-based services. - banking reform must include continued interest rate reform, recapitalization, and establishment of lending guidelines and a regulatory system controlled by the Central Bank. - equity financing for industrial rehabilitation should be mobilized. probably by establishment of a specialized investment company 1.04 A principal aim of the study is to explain and elaborate key aspects of an enabling environment for business, and for industry in particular, in Mozambique. An enabling business environment is defined as one that permits business to operate within a manageable and comprehensible institutional and regulatory system, and with a supportive macroeconomic and incentive framework. This is the first effort to draw together the wide range of disparate issues and policy areas which are instrumental in defining the business environment in Mozambique and it is inevitably selective. 1.05 The idea that the central issue of industrial strategy in a developing country should be the creation of an enabling environment is quite different from the conception of the strategies for industrial development which were formulated in the post-independence period of the 19609 and 1970, in many African countries. The latter traditional" approaches were more concerned with establishing what was conceived as the industrial base. This involved in many African (and other) countries direct state involvement in what were identified as key industries, particularly essential consumer goods and heavy industries, and a concomitant demphasis on the private sector. In particular, private foreign investment and ownership was in many cases considered to be inconsistent with the development of an appropriate industrial base, inter alia because of its supposedly short-term horizon, its perceived monopolistic tendencies and its reluctance to reinvest profits and transfer technology within host countries. To support the state-sponsored industrialization effort it was necessary to develop: (i) an investment and production planning system, often on a five-year basis, and; (ii) a trade protection and internal regulatory system which would permit infant public enterprises to operate profitably. The formulation of an incentive system for private business was of secondary importance, and the international competitiveness of local industry was subordinated to the principal concern of choosing priority industries and setting up pro'uction plants. The competitiveness of industry was assumed to be achievable as a result of a learning process which would reduce infant industry costs over time. 1.06 The experience of the past twenty years has shown that, despite some initial gains in terms of industrial growth, usually of an import- substitating type, the state planning and investment approach has not been succassful in many countries, and, rather than laying an industrial base, it has resulted in many cases in economically inefficient, high cost, capital and import-intensive enterprises, often unprofitable without sustained protection, and highly vulnerable to foreign exchange shortage without contributing adequately to foreign exchange savings. In fact the decade of the 1980s has been one of deindustrialization in parts of Africa; the earl!er gains in industrial growth have been reversed so that in 1987 the ratio of manufacturing tc GDP in the lower income economies of SubSaharan Africa was the same as it had been in 1965, at the start of the industrialization drive. 2/ 1.07 As stated above, an enabling environment is one that involves a combination of sound macroeconomic policy, price incentives, and a supportive institutional and regulatory framework. More precisely the business environment msy be defined inter alia by: (i) the nature of the price system - the exchange rate, interest rate, wage rates, and input and output price determination; (ii) the trade policy regime - import and export tariffs, subsidies, trade-related goods taxes, and quantitative restrictions, including licensing regulations etc; (iii) the corporate tax system and the system of fiscal incentives; (iv) the financial system - its solvency, liquidity, lending regulations, supervision, capacity to provide industrial finance (debt and equity), and system of monetnry incentives; (v) the role of the state, ownership and reservation policy for industrial assets and land; (vi) corporate and labor law and its enforcement, particularly as regards entry and exit regulations for industry; (vii) the institutional and economic infrastructure (promotion and technical services, power, transport etc). An improvement in the business environment, in the context of a controlled economy with nonfunctioning distorted markets, involves freeing of prices accompanied by removal of quantitative restrictions in foreign exchange, capital, labor and goods market. Physical controls are replaced where necessary and appropriate by transparent price-related measures (e.g. taxes and tariffs rather than licensing controls). Producers are then able to make autonomous buying and selling decisions based on rational price signals. Reforms in the regulatory and incentive system should complement price reforms, aiming at encouraging efficient investment and production. 1.08 Over the past three years, since the announcement of the Economic Rehabilitation Program, a series of policy initiatives has affected the business environment in Mozambique, in the area of exchange rate management, planning, pricing, distribution, public expenditure, credit, investment 2/ This is discussed in SubSaharan Africa: From Crisis to Sustainable Growth, The World Bank, Washington, 1989. -4- incentives, enterprise avnership. These initiatives have entailed a broad series of changes, notablys (a) the opening up of access to f.: .gn exchange, with improvements in competitiveness of industrial production through exchange rate changes, and modification of the tariff structure to substitute for quantitative controls, and to reflect more the requirements of industry for protection; (b) an increase in enterprise management autonomy and accountability after 10 years of central planning of pricing, distribution and procurement of larger e- erprises with few major decisions taken by management; tc) improvement in financial discipline as a result of the cessation of the use of the banks as a channel for financing the central plan, credit and interest rate reform, redefinition of the role of the Central Bank, restructuring of bank finances, and decentralization of lending; (d) some revival of the private sector through changes in ownership policy, investmen incentives, increased interest by foreign and local investors, and increased emphasis on the development of swall enterprises. 1.09 At the same time great efforts are still needed to inter alia rationalize enterprise ownership, rehabilitate and increase utilization of capacity and improve management, as well as continue exchange rate, price and wage reform, and rehabilitation of the banking system. As the central planning of production, prices and distribution is dismantled, the role of Government in industry has to become one of more regulation and oversight. This requires creation of the right conditions for industry to flourish, and implies the need to construct an incentive framework which channels production and investment, both in skills and fixed assets, in ways that are consistent with economic priorities. 1.010 The study examines the following specific policy areas relating to medium term industrial strategyt (P) industrial structure and growth and short to medium tern industrial efficiency in terms of international competitiveness; (b) the efficient allocation of resources and appropriate policy in the areas of foreign exchange allocation, licencing, prices, trade tariffs and taxes, corporate taxation and fiscal incentives; (c) analysis of the financial sector, the reform process and future developments, and aspects of industrial investment policy, including the export retention scheme, asset revaluation, debt-equity conversion, and promotional infrastructure; -5- (d) t.e requirements for industrial investment financing and possible instruments for private sector participation, with emphasis on equity financing, and their role and effectiveness in industrial promotion. 1.11 The enterprise data-base for the study was a sample of about 60 enterprises visited in Maputo, Beira, and Nampula, and 10 enterprises included in a detailed enterprise restructuring study. 3/ In addition a credit survey of 120 companies was used as the basis for estimating demand for industrial finance. B. Conclusions and Recommendations Industrial Structure 1.12 The industrial sector in Mozambique inherited from the colonial regime was relatively diverse, and large in terms of output per capita, by comparison with other Subsaharan African countries; there are few obvious instances of misallocation of past investment, most of which took place prior to 1975, during a period when Portugal opened the economy to multinational ivestment; potentially many enterprises could probably be rehabilitated to =come economically efficient. However, industry as a whole is import- Spendent and has an ageing capital stock some of which was second hand at istallation; the state of plant and equipment in many enterprises is very )or and is a constraint on efficient production. However, over 1987-89 _idustrial production has increased following a steep decline from 1981 to 1986. The increases are due mainly to increased provision of materials and spare parts through donor support, allowing fuller capacity utilization, and initial steps in rehabilitation of enterprises. Donor support however brings with it problems of production planning, procurement tying, and lack of coordination. Effective further recovery depends on major rehabilitation where feasible, but it may also face a depressed market for some manufactures, due to the decline in real purchasing power of certain sections of the population, and dislocation of others as a result of the security situation. Manufacturers of discretionary consumer goods have faced market constraints. 1.13 The general trend of declining output has been countered in the case of the micro-enterprise sector, those of 10 workers or less, which, according to available data have shown an increase in production and employment over 1973-1987, especially in metalwork, wood products and garments industries. Thia sector, if economically efficient, may also be a source of future growth, and it is also likely to be a source of entrepreneurial learning. A focus of future credit and promotional activity ihould be in this sector, in conjunction with the effort to rehabilitate existing larger enterprises which account for the majority of employment and capacity. 3/ This study was carried out for the IDA Industrial Enterprise Restructuring Project. -6- 1.14 Tht-re are indications of increased local and foreign investor interest in the country in a number of sectors of industry and commerce. The Government has started to improve investment incentives with reforms of the tax and tariff system and provision of investment codes, ationalization of company ownership, and financial restructuring of enterprises and the banking system. In addition major macroeconomic reforms such as currency realignment have had beneficial effects on industrial competitiveness. However, the prospects for both foreign and local invet-ment depend to a significant extent on the resolution of the security situation. Economic Efficiency of Industries 1.15 Currency devaluation has generally lowered wage costs as a proportion of total costs so that enterprises are more competitive in terms of international prices. In the short run, taking into account direct costs and ignoring plant replacement costs, most industries are able to produce efficiently in economic terms. This means that they are generating net foreign exchange and doing this efficiently enough to justify the commitment of resources for short-term support. Enterprises which are not economic in the short run appear to include sugar, glass, tire tubes, and metal pipes. which are not covering their direct operating costs in the present difficult operating conditions. 1.16 Taking account of long term costs of production including plant replacement costs, with their present setup many industries are probably uneconomic. This is due to (a) shortage of materials, inadequate repair and maintenance and lack of technical and management skills, which lead to capacity underutilization, (b) operating inefficiencies such as overmanning and excessive material wastage and (c) demand constraints resulting from dislocation the population and economic adjustment. However, rehabilitation of these industries, and the sector in general, with complementary increases in productivity in the workplace and good management, could in principle restore economic profitability to many enterprises in the long run. The rehabilitation of industry will firstly require identification of industries which are economically viable, and, secondly, improvements in labor and material productivity in those industries. Case-by case cost-benefit studies are essential if a viable restructuring program is to materialize. Industrial Sector Policy 1.16 Foreign Exchange Allocation and Licensing. The current export retention scheme is an effective short to medium-term instrument for mobilizing exports, although it may create certain longer term investment and production biases. The operation of the scheme could be clarified in order reassure investors that they do not face significant foreign exchange risk. In this respect some benefit may be derived by setting up special overseas accounts for eligible companies which would provide guaranteed access to foreign exchange, unaffected by exchange rate changes. 1.17 Prior to the opening up of the foreign exchange market, the Government should institute a foreign exchange allocation system based on a measure of the efficiency of production of enterprises, as a substitute for the current more arbitrary priority lists. The suggested measure is the DRC ratio which is already known to the authorities. A method for systematic prioritization of foreign exchange applications from enterprises using efficiency criteria should be introduced. 1.18 Prior to the opening up of the foreign exchange market, the policy of non-issuance of import licenses when domestic suppliers are available should be amended in order to ensure that lecal products are not isolated from international prices. In order to achieve this a specific percentage margin of domestic price preference for local suppliers could be introduced. This would assist the transition towards the use of tariff and exchange rate instruments to control foreign exchange demand, which is preferable to licencing. 1.19 Trade Tariffs and Pricing. By international standards tax revenue in Mozambique is relatively highly dependent on domestic taxation, with trade taxes comprising a relatively low proportion. Quantitative controls are used to restrict imports. However such controls do not produce tax revenue but instead transfer monopoly rents to traders. In addition controls obscure the relationship of domestic to world prices. The use of tariffs would restore the sensitivity of domestic prices to changes in world prices, and is therefore preferable on trade efficiency gremnds. In addition it would boost Government revenue. The tariff system should be designed for a) uniformity, and b) transparency. The study recommends the following. 1,20 A tariff commission should be formed within Government to monitor the issue of protection of industry (and other sectors) and advise on changes. The role of the CoDmission will be to: (a) estimate and set a minimum uniform tariff on all goods based on trade and exchange rate considerations, and revenue needs; (b) identify nonessential consumer goods which qualify for either higher tariffs or special excise taxes, on distributional equity grounds; (c) identify possible industries which qualify for infant industry/rehabilitation protection or protection in terms of employment or expected long run productivity gains; (d) examine the possibility of setting up an indirect duty drawback to exporters to compensate for high cost tariff protected domestic inputs. 1.21 In order to move towards a rationalized system, of protection, capital equipment and industrial materials should not be exempt from tariffs. 1.22 Prior to general decontrol of prices, and assuming that demand will sustain the prices set, the conditional pricing system should be modified and based on import price plus a margin, rather than the existing cast-plus- markup formula. This is to ensure sensitivity of prices to international competition. The margin proposed would be linked to that for 'domestic preference'. Enterprises setting prices within the agreed margin would, as at present, set their own prices with reference to the authorities ex-post. - 8 - 1.23 Corporate Taxation and Investment Incentives. Tax holidays should be generally restricted in length. However, flexibility in negotiating with investors is important in order to maintain the current momentum of investor interest. Profits tax should be further reduced over time and the depreciation allowance incentives should be curtailed. 1.24 The Government should start looking into ways of substituting the Imposto de Circulacao (turnover tax), which currently tends to operate in favor of imports, by an alternative such as a value added tax or consumption tax. 1.25 The access of small-scale enterprise to fiscal incentives may be restricted because of lack of information; it is recommended that a study is made to letermine whether this is a constraint on development. 1.26 The above combination of tax proposals would increase the present value of Government revenue while restructuring the system of incentives to give greater incentive to existing producers, and maintaining a satisfactory level of new investor incentives. Financial Sector Policy 1.27 The financial system, and industrial investment promotion. There is considerable unsatisfied demand for long term rehabilitation finance from potentially economically viable enterprises. In order to effectively meet this demand continuing structural reforms in the financial sector are desirable. (a) The Banking system requires reorganization in order to strengthen its competitiveness and operations, and to increase productive savings, investment and growth. This involves inter alia institutional diversification (including redefinition of the role of the Central Bank), clearing of lending arrears, recapitalization, and reestablishment of bank performance criteria and a regulatory framework for commercial bank operations. The revaluation of enterprise assets, currently under way, is an essential requirement for the revival of sound commercial lending as well as regularization of company taxation. (b) The foreign investment legislation is in general beneficial to new investors, although some refinements may be advisable. General clarification of corporate law is important, p&rticularly in areas such as ownership and transfer of companies, industry entry and exit, and labor legislation. The debt-equity swap law appears relatively restrictive, especially concerning reversion of companies to the state and remittance abroad of principal. Clarification may be necessary if this scheme is to be effective. (c) If the former Unidades de Direccao are reconstituted as holding companies, it is important to establish them as self- financing entities independent of Central Government, and to remove certain functions, such as foreign exchange allocation, in order to prevent conflicts of interest. (d) In order to maximize effectiveness of small enterprise promotion, the Government, through IDIL and other appropriate agencies, should ensure a) clear identification of needs of small businesses, b) cost effectiveness of assistance measures, and c) cost recovery through generation of revenue, where possible. 1.28 Financial Instruments and the Mobilization of Equity Funds. The rehabilitation of industry requires equity as well as debt financing. An environment more conducive to equity ownership should be developed, encouraging resources of entrepreneurship to come forward. The Government should examine institutional options in this area, building up on existing experience in the banking system and new institutions such as GPIE, and SOC7EF. Financial instruments should be introduced to permit transfer of control of state or intervened companies while, where appropriate, preserving the essential interests of the Government. Among a number of available financing techniques, appropriate instruments may include nonvoting preferred shares, and a form of management agreement with enterprise investors, which would allow independence to investors while safeguarding Government shareholdings. 1.29 For enterprises which have been rehabilitated, the possibility of worker share participation schemes should also be examined, as an incentive to higher productivity and as a redistribution device, building on the experience of generally constructive worker- management relations in Mozambique since independence. The success of such schemes however depends on the implementation of reforms in the business environment. 1.30 Considering the large number of industrial enterprises currently under state control, the Government should examine the possibility of forming a type of investment company, possibly with its own capital, which would mobilize equity investment, and other sources of finance, for industrial rehabilitation, and would provide services in the preparation and packaging of company restructurings, transfers and new investments. Initial financing of such an institution could probably be identified from sources such as international donors, and the private sector. -10 - II. ECONOMIC BACKGROUND AND INWUSTRIAL GROWNH A. General Economic Backgron 2.01 At independence the economy of Moz..smbique was built around four main activities - subsistence agriculture, processing of agricultural commodities for export, migrant labor, and transit/tourism services. The external account depended on merchandise exports (principally cashew nuts, prawns, cotton and tea), transit trade (from South Africa, Zimbabwe and Malawi) and tourism, and mino workers remittances, (largely from the South African coal mines). The ue from remittances and services was usually sufficient to offset its on the merchandise trade account, which ranged between 202 and 352 bport value over 1960 to 1973. The industrial sector expanded rapidly a low base in the 10 years preceding independence, following the opening 'the economy to private foreign investment. Because of the relative ess of the economy over 1965 - 70 new industries needed to be --tionally competitive and therefore the sector does not exhibit obvious cation of investment; however, much equipment was second hand and all ent and skilled labor was foreign. The increase in industrial ------ent also assisted the build up of external debt, which by 1974 was US$640 million, or approximately 702 of GDP. 2.02 Following independence a series of major setbacks occurred. These included the departure of 902 of the settler population (about 200,000 people), and virtually all skilled manpower over 1974 -77, which obliged the Goveranment to take over control of numerous industrial enterprises in order to maintain production. The emigration also had particularly severe effects on plantation agriculture, rural distribution and the railways and ports, In 1974 foreign exchange reserves were almost fully depleted. The overall situation was aggravated by the diversion of transit trade by South Africa, reductions in mine labor and remittances, and fall off in tourism earnings. 2.03 Apart from the problems inherited at independence, in 1980 serious difficulties arose with the breakdown of the security situation; combined with a major flood and drought this led to a 602 reduction in exports over 1980-84. This was compounded by deepening economic management problems resulting from the central control of prices and production, which resulted in major exchange rate and price distortions in the production and financial sectors, with diversion of goods on to parallel markets. In 1986 the parallel market price for foreign exchange rose to a level twenty times the official rate. The shrinking of the tax base which had resulted from the dejarture of the settlers, and the price controls with consequent operating losses of many enterprises, assisted the growth of the fiscal deficit which reached a peak of 17.72 of GDP (before grants) in 1986. Concurrently, by 1986, total external debt reached US$3.0 billion and the debt service requirements exceeded 2002 of exports of goods and services, largely bilateral public debt, of which 202 was in arrears. Since 1987 the balance of payments gap has been met by donor assistance amounting to 80-852 of Mozambique's foreign exchange inflows. As a result of the fundamental problems facing the economy GDP per capita is estimated to have declined by 82 per annum in real terms between 1973 and 1983. - 11 - 2.04 The industrial sector contribution to GDP rose to a peak of 12.5Z in 1980 but fell steadily over the period since. The decline in the services sector (which was the largest contributor to GDP in 1973) paralleled industry up to 1984, reflecting the collapse of transit trade and tourism. The manufacturing contribution to rDP in the years 1973 and 1980, was greater than of many other Sub Saharan African economies; however the 1985 level was well below comparable economies. reflecting serious disruption of industrial output and high level of excess capacity. 2.05 In January 1987 an Economic Rehabilitation Program was initiated. Overall details of the ERP are set out in Government documents such as the Policy Framework Paper, and its relevance to industry and banking is discussed in the remainder of this report. In overall terms the ERP introduced major macroeconomic changes designed to address the serious economic problems facing the country. Economic stabilization measures centered on credit controls and public expenditure cuts, while structural adjustment measures focussed on the exchange rate, pricing, trade and distribution of goods, taxes and tariffs, and a series of other measures related to these. The state Central Plan, in force since 1978, has been reduced in scope to more of a review function. As a result of the ERP measures, and major international financial support, including foreign debt service renegotiation, real GDP is estimated to have risen by about 42 per annum over 1987 and 1988. 2.06 The major source of growth has been the family and private commercial farm sector which responded to improved price incentives and probably also to the increased availability of consumer (incentive) goods. In addition industrial production turned round in 1987 and has expanded by over 252 up to 1989. A recovery of 302 in merchandise exports occurred from the low point of US$79.0 million in 1986 to an estimated US$103 million in 1988. Inflation and fiscal deficit targets have been met, the exchange rate target has been partly met and has settled at around 40 to 502 of the parallel rate, implying a considerable realignment of internal and external prices. The impact of the ERP has to be seen in context however; production levels still remain well below 1980 levels and the serious problems of security remain, along with those of inadequate skilled manpower and rundown capital stock. B. The Development of the Industrial Sector Industrial Developments Pre-independence 2.07 The current industrial situation in Mozambique is the outcome of ,everal decades of relatively drastic political and economic change, alongside changes that were also occurring in Portugal. In the early 20th century, industry was confined to a few agroprocessing plants such as sugar and sisal; manufacturing growth originated at the start of the industrialization of Portugal itself, in the 1930's, when Portugal turned its attention to its colony as a source of raw materials. From about 1930 smaller scale manufacturing began to emerge in Mozambique - including cement, bricks, -12 - soap, beer and cigarettes, and a market for these industries was provided partly by the settler community which at that time numbered about 20,000. At the same time certain industries were curtailed in the colony, in order to ensure that markets were open to Portuguese exporters. This was the case with the Portuguese textile industry, based on Hozambican cotton exports which rose fourfold between 1939 and 1946. 2.08 Between 1946 and 1961 limited industrlal growth occurred largely to supply consumer goods, and to provide inputs to the export processing sector. Total employment grew from 25,000 to 90,000. The emerging industrial sector diversified from primary processing into consumer and intermediate manufacturing, including food products, garments, footwear, furniture, glass, metal products and light engineering. The first textile factory was permitted to start production in 1952, and cement production rose from 26,000 tons in 1946 to 200,000 tons by 1960. Nevertheless the most important subsector (in terms of employment and investment) remained primary processing; i.e. tea, sugar, sisal fibr, cotton ginning, and edible oil. 2.09 Over the period 1957 to 1970, according to the available index of industrial production, the rate of growth of industrial output averaged between 6 and 72 per annum, with an acceleration to over 102 during the period 1965 -69, as Portugal opened up the economy to foreign investment, particularly from South Africa. The 'open door' polic. for foreign investment included a series of incentives such as tax holidays, a. 1 profit repatriation entitlement; previous regulations requiring financial participation by Portuguese firms were dropped. In one sense the legacy of this period has been helpful to industrial viability, because the open door policy forced new enterprises to be internationally competitive and would therefore have resulted in the establishment of economically efficient operations. However, these enterprises tended t3 be highly dependent on imports, and were often equipped with second hand machinery; their reliance on foreign management and technicians was also a potential weakness. 2.10 The domestic market was boosted by a surge in colonization, with the number of settlers rising to about 240,000 by 1970, largely residing in the urban centers. The most rapid growth by subsector was in the metals and engineering industry which recorded an increase of 30? per annum in the latter period (1964-70). The restrictions on the textile industry were lifted, although only one additional plant had opened by independence. Intermediate and capital goods production expanded, including engineering, and metals, a petroleum refinery, chemicals, fertilizer, paints, etc. Import substitution manufacture increased its share of total industrial output steadily, from approximately 37? in 1955 to 61? in 1972, while the share of export processing activities declined proportionately. There was also extensive South African investment in cashew nuts with backward linkages to tin can manufacture and gas supply, both servicing cashew exports. By 1973 Cashew processing comprised about 25? of organized industrial employment (over 20,000 workers), and the industry was a major world supplier. 2.11 Summary data on enterprises and employment in manufacturing and processing over the period prior to independence in 1975 are as follows. - 13 - Table 2. 1: MOZAHIQUE - PREINDEPENDENCE GROWTH OF ENTERPRISES AND EMPLOYHENT 1962 1967 1969 1970 1972 1973 No of enterprises 1283 1602 1802 1904 1488 1418 Employment 59090 68653 73044 85050 95810 99503 Sources Estatisticas Industriais; various years; Instituto Nacional de Estatistica 1973. In 1973 total manufacturing value-added per capita in Mozambique was estimated to be the sixth largest in SubSaharan Africa. 1/ The share of manufacturing in GDP was also higher than that of many Subsaharan Countries. Industrial Decline: Post-independence 2.12 Following independence in 1'75, the large-scale exodus of Portuguese settlers over the 3 years 1974-77, caused a serious disruption in production. The settler community had all but monopolized managerial and skilled jobs, largely because of the lack of educational facilities for Africans. This situation necessitated the "intervention" by the Government in numerous enterprises, many of which were by then in a state of neglect, and in some cases with heavy debt arrears. The coal mines were taken over after 2 serious accidents. Formal nationalization however only occurred in the case of petroleum refining and a shipping company. The banks (except for Banco Standard Totta which remained private) were either closed or merged under state control. The state also set up a network of overseas and domestic trading companies to take over commercial operations necessitated by closure of Portuguese firms. In 1975 specific criteria were laid down for intervention. These included a 60-day delay following the publication of a statutory notice during which owners of abandoned units or their representatives could oppose. In 1977 it was further decreed that intervened enterprises could be taken into state ownership if necessary, subject to various provisos. A total of approximately 300 enterprises were intervened initially, including many small distribution outlets which were after 1980 returned to the private sector. In 1987 a total of 140 manufacturing and processing enterprises were listed as intervened. It appears that the Government has returned numerous enterprises to private ownership, mainly in the small scale distribution sector, but increasingly in manufacturing. 2/ 1/ See UNIDO Handbook of Industrial Statistics, UN, 1984, Table 1. 2/ A notable example is Habor Tyres, where the Government renegotiated agreements with General Tyre. Similar initiatives have been taken in other csinponies. - 14 - 2.13 A partial index of industrial Vioduction for 1973-Z9 is shown in figure 2.1. Figure 2.1 INDEX OF INDUSTRIAL PRODUCTION P0VISILGA C1SS0 * 100) 140 130 - 120 110 100~ soX_ j 70. 60 -- 50 20 10 1973 1975 1980 198 1984 199 1987 1988 1989 YEAR 2.14 The index of production is tentative. It excludes certain subsectors, and applies to nationally registered enterprises (approximately 50 workers and above). Estimates for 1987 and 1988 are based on reported year to year increases. 1989 projection is based on MIE figures only. 2.15 The rapid decline in output after independence which is shown in the table was partly reversed by 1980. In 1981 however sabotage of infrastructure by armed bands began. This was combined with external economic difficulties associated with the world recession, and the lack of skills and management capacity. The centralized control of production, distribution and pricing also prevented enterprises adapting to adverse conditions. As a result there was a steep decline in industrial output over 1981-85 with 1986 output at less than half its 1V80 level for the above group of industries, and 30X of the pre-independence peak production. Sabotage was responsible for the decline of electric power distribution from 11.3 GWH in 1980 to 0.4 GWH in 1986. Petroleum refining was halted in 1985 and major output reductions occurred in a number of other manufactures, as shown in table 2.2. - 15 - 2.16 As a result of donor import-financing and the start of rehabilitation, a turnaround in production occurred in 1987 when output rose in real terms by 7.0X; this was followed by an estimated 9.7Z increase in 1988, and about 7.6Z in 1989 according to partial results. The extent of disruption of industrial production is shown by the fact that, despite the three years of expansion, 1989 estimated output would still be little more than 5O0 of tha; achieved in 1980, and below 40? of 1973 levels. Over 1985- 88 output recovered irn a wide range of industries, but continued to decline in clothing. leather, footwear, furniture, and metals. 2.17 The steep decline in output volume in principal subsectors is shown in table 2.2. Table 2.2: MOZAMBIQUE - OUTPUT VOLUMES: PONER AND PRINCIPAL MANUFACTURES 1980 1981 1982 1983 1984 1985 1986 1987 Electricity GWHOOO 11.1 3.6 3.0 6.0 .4 (J.5 0.4 0.5 Textiles (mn mts) 6.0 10.1 9.5 8.4 8.8 6.1 5.2 7.2 Garments (mn pcs) 7.6 0.8 7.1 8.9 0.2 6.7 4.7 4.8 Paints (mn liters) 1.0 0.6 1.0 0.9 0.7 0.4 0.2 0.3 Tyres,tubes (000) 197.0 301.0 228.0 195.0 89.0 162.0 123.0 a/ Glass cont (m pcs) 24.0 121.9 19.9 16.5 20.6 9.3 6.9 a/ Oils,Soap (000 tn) 36.8 47.5 40.6 34.1 21.4 12.6 13.2 17.3 Flour (000 tons) 98.4 109.5 95.0 82.0 74.4 73.6 55.5 89.1 Cement (000 tons) 236.0 261.0 270.0 188.0 105.0 77.0 73.2 73.1 Timber (000 cu mt) 164.0 140.0 92.0 62.0 74.0 68.0 72.0 a/ al Statistics -were reclassified in 1987, and these categories are not directly comparable with 1986 data. Sources Estimated from Informacao Estatistica 1986, 87 (Conmmissao Nacional de Plano), 2.18 In the case of agroprocessing even larger declines occurred in marketed production, exacerbated by demonetization resulting from the general shortage of consumer goods in rural areas. In 1987, 6 ort of 14 cashew processing plants were non-operational, either closed down or sabotaged, according to the FAO, 31 with output at 15-20Z of 1973 levels. The industry has been decimated by a combination of a) neglect of plantations, b) drought, fire and pests, c) breakdown of the grading system in the field d) poor product quality and e) lack of transport vehicles, and attacks on rail freight. 2.19 Capacity utilization of industrial plant is estimated to have fallen to less than 30? in 1986 in many industries. (However the measurement of utilization is problematic). The most seriously affected industries were sugar (62 utilization), cashew, and oils and soaps (10Z) because of their '$/ Sources Study of the cashew nut industry in Mozambique. FAO, 1987. - 16 - reliance on domestic materials and transport subject to sabotage. In manufacturing the most seriously affected were containers and paints (both estimated at 8Z), while only batteries and radio assembly recorded higher than 50. 2.20 Despite the dramatic decline in industrial production, industrial investment since independence took place to a limited extent in cereals processing, ship repair, bus assembly, timber processing, farm implements, vegetable oil, and textiles. In 1987 a total ef 20 textile-related projects operated in the formal sector, emplopying 12,000 workers. Industrial investment financed by the state reached a peak in nominal and real terms of 251 of total public investment expenditure over the period 1981-83, averaging about Mt 4.0 bn per annum. However, since then there has been a decline to the 1986 level of Mt 1.30 bn, accounting for just over 10 of public investment which was itself well below the 1981-83 levels. Over 1987-89 public investment in infrastructure and health services has increased, with the proportion of investment in industry and energy declining to a projected 61 in 1990. Manufacturing Employment 2.21 Industrial employment suffered along with output, although labor laws over the period prevented a steep reduction in factory manning. There have been a number of surveys of employment in industry, leading to conflicting estimates. Table 2.3 sets out the probable levels over time. Table 2.3: MOZAMBIQUE - GROWTH OF MANUFACTURING DIPLOYMENT 1973 a/ 1984 bl 1988 c/ 1988 d/ (revised) Total 99,868 83,533 116,989 88,553 a/ Source: Estatiaticais Industriais 1973: all enterprises, with possible underreporting of microenterprises. h/ Source: UNIDO: (1987) p 56t national level enterprises only. c/ Sources data supplied by Commissao Nacional de Plano (CNP): national level enterprises plus a proportion of local enterprises. d/ Source: amended CNP data. 2.22 The 1988 estimate is adjusted to exclude sugar and cashew estate workers which were included in the manufacturing employment data supplied by CNP. A survey by the Ministry of Labor for 1987 estimated 66,500 for all manufacturing, including microenterprises, which is inconsistent with the UNIDO and CNP estimates based on only national enterprises. The Ministry of Labor estimatii probably underreported employment outside Maputo. From the statistics avdilable it appears that there was a significant fall in employment, but not of the scale of the output decline. - 17 - Exports of Processed Goods 2.23 Exports suffered from the decline in production of traditional products, particularly sugar and cashew nuts, whose export volumes in 1988 were 12.5Z and 41.7Z, respectively, of the 1980 level. The composition of processed export products has been primarily the following: cashew kernels and oil, ginned cotton, sugar and molasses, and tea. In addition, in the past occasionial exports have occurred of cement, tires, fabricated metal products and other manufactured goods. This is shown in table 2.4. Table 2.4: MOZWMBIQUS - PROCESSED EXPORT PRODUCTS 1973-87 Year (US$ million) Year (US$million) 1973 138.4 1983 79.0 1975 93.2 1984 47.1 1980 202.8 i985 33.4 1981 184.4 1986 42.4 1982 148.6 1987 50.6 Source: Informacao Estatisticais, various years. C. The Structure of Industry The Composition of Industrial Output 2.24 The following table shows the structure of employment and value added by 2 digit ISIC subsector, for 1973 and 1988 over all manufacturing and processing. Table 2.5: MOZAMBIQUE - STRUCTURE OF MANUFACTURING OUTPUT (Subsector Shares at Domestic Prices) 1973 1988 ISIC code Employment Value Employment Value added added . A . Food products 311-312 42.1 29.6 35.9 20.7 Beverages, tobacco 313-314 4.2 14.3 2.8 20.8 Textiles, garments, 32 leather, fibre 16.9 15.8 26.3 23.0 Wood, paper, printing 33 -34 16.0 10.3 11.6 8.2 Petroleum, chemicals, rubber, plastics 35 4.1 8.8 5.7 10.7 Nonmetal. mineral prods. 36 4.7 7.0 5.4 4.5 Basic Metals 37 1.1 1.2 1.9 1.2 Metal products, equipment 38 10.5 12.3 9.9 10.8 Other industries 39 0.4 0.5 0.4 0.1 Total 100.0 100.0 100.0 100.0 Sources Based on Estatistical Industriaw.s, 1973 , Ministry of Labor, 1987, and mission estimates. - 18 - 2.25 The estimate of value-added for 1988 is provisional, based on mission estimates. 4/ Subsector shares measured at World prices would differ from those at market prices. 5/ Despite the externally dependent situation of the economy (in terms of trade, technolog> and finance), in 1973 the level of industrial capacity was relatively significant. As stated, manufacturing value-added per capita was the sixth largest in Subsaharan Africa. Value- added per capita for three major industrial branches was as follows: Table 2.6* MOZAMBIQUE - LEVEL OF MANUFACTURING VALUE-ADDED (1973): SELECTED LOW-INCOME AFRICAN ECONOMIES (Per capiti in US$ at 1975 prices) Total Food, drink Textile, garments Metal products Tobacco Leather, footwear Equipment (31) (32) (38) Ethiopia 11 3 4 0.2 Kenya 29 8 3 7 Madagascar 29 7 12 3 Tanzania 18 5 4 2 Zaire 15 3 3 3 Zambia 86 9 10 27 Zimbabwe 138 17 22 33 Mozambique 48 21 7 6 Source: UWYDO (1984); Op Cit, and Estatisticas Industriais. 4/ Value-added statistics have not been officially compiled. For this study estimates of manufacturing value added were made by applying separate value-added to gross output ratios for each three-digit branch of manufacturing, based on international data and derived by estimating the average ratio for individual branches for a sample of developing countries. Source: Industry in a changing World UN, New York, 1984, p.215. S/ Measured at world prices, the structure of output would differ if effective rates of protection are variable across branches. Shares of domestic resource-intensive industries (foods, textiles, wood, building materials etc), would tend to rise relative to import-dependent industries (petroleum, chemicals, plastics, rubber) which are generally more highly protected. Effective protection is difficult to measure in Mozambique because of extensive trade restrictions. (See chapter 5.) - 19 - 2.26 In 1973, in terms of industrial structure, the dominant subsector was food, drink, and tobaccc, implying a colonial economic orientation (i.e. dominated by the processing of agricultural raw materials). The largest single industry in terms of value added contribution was sugar, and the largest in terms of employment was cashew processing. The dominance of foods resulted in relatively modest shares for textiles, metals and other branches. However the metal products and equipment subsector was larger in absolute per capita terms than that of four other countries illustrated, aithough considerably lower than that of Zambia and Zimbabwe. The textile industry group was also larger in terms of value-added per capita than that of several other countries. The split between consumer, and producer (intermediate and capital) goods production was approximately 67 s 33 in terms of value added. The direct export sector comprised about 20X. Overall average scale of production was 70 workers per establishment. Despite the relatively high level of industrial activity in relation to population, there were built-in problems, par%. -larly because a proportion of plant was second hand when installed, and tne skilled labor force to maintain it and operate industrial enterprises was almost exclusively foreign. 2.27 The 1988 estimates show that the structure of industrial production has shifted significantly towards the textile and apparel subsector, with a major decline in the case of food products; in the textile industry there was a high rate of investment in the 1970s, and as a result absolute declines in production were lower than in other industries; in the case of foods the decline has been mainly due to low capacity utilization and deterioration of assets because of security-related problems. This also applies to nonmetallic mineral products, in particular the cement industry which is still producing at less than 30Z of its 1981 level. In the case of metal products and equipment the decline has been more due to technical problems (lack of skilled manpower) and scarcity of imported inputs. The 1973 structure of output may be closer to that which would obtain if enterprises were currently operating nearer their capacities. Industrial Location 2.28 In 1973 Haputo (Lourenco Marques) province accounted for 442 of employment, and 51? of industrial value added. Beira and Maputo together accounted for 60Z of employment, and 71? of value added. The most widely distributed industries were cereals milling, cotton ginning, sisal, and timber which were located largely outside Lourenco Harques/Maputo province. In general the food and fibre processing industries, as expected, were widely distributed, while manufactured goods such as rubber, plastics, chemicals, paints, and metal products were heavily concentrated in the two major towns. 2.29 In 1987 the pattern was altered by the considerable increase in the share of Maputo province. This may partly reflect underreporting elsewhere. However, absolute employment in manufacturing and processing in Haputo Province, according to the Ministry of Labor Survey, shows an increase from 43,000 in 1973 to 46,400 in 1987. Maputo and Beira continue to dominate the sector in terms of employment and output. There are advantages in this - 20 - in terms of economies of concentration and security considerations for those enterprises not dependent on materials from upcountry. However, against this in the longer term is the issue of income distribution and regional balance. Ownership Structure of Industry 2.30 As stated, following independence the ownership of industry ehanged radically as a result of the decision by the state in 1975 to tervenel abandoned or commercially sabotaged enterprises. After an tial intervention in about 300 enterprises the 1977 law permitted various nges, including full state ownership, or reversion to private ownership, ending on the circumstances. (See para 2.12.) In 1984 the structure of uw..ership of 575 nationally registered enterprises was as follows: Table 2.7s 7 HOZBIQUE - OWNERSHIP STRUCTURE OF NATIONAL ENTERPRISES state 114 intervened 140 private 294 mixed and cooperative 27 total enterprises 575 Source: UNIDO (1987) 2.31 State control over the manufacturing sector, partly forced on the Government by circumstances, is relatively extensive, and possible options for divestment are under continuous consideration. Evaluation and Sales Committees were set up following the law of May 1989, responsible for overseeing the disposal of intervened assets and the Government recently announced its intention to establish a bureau of enterprise divestment. This is discussed further in chapter 3. D. Industrial Strategy and Plannings 1977 - 87 2.32 In the period immediately after independence strategic issues had to be shelved as the Government concentrated on short term measures to offset the effects of the departure of skilled manpower. By 1980 however, prospects had improved, with partial recovery in production and favorable political developments in Southern Africa. As a result the Government decided to launch a 10 year development plan for 1981 - 90, with the aim of eradicating underdevelopment in a decade. The plan was extremely ambitious, calling for 17Z per annum increases in GDP, with parallel large increases in all sectors, with emphasis on large scale projects, state farms, and heavy industry, including iron and steel, aluminum, natural gas, chemicals and cement, plus expansion of consumer goods such as textiles. -21 - 2.33 The financing and implementation demands of the plan however forced its abandonment, prior to formal publication and it was substituted by a more modest three year plan for 1982 -85. This in turn was upset by the rapidly developing economic difficulties of the country, and in 1983 at the 4th party congress a new policy was adopted emphasizing the increase in utilization of existing capacity, rehabilitation, and acceleration of existing project implementation. The previous emphasis on large projects was also substituted by support of small scale industry and liberalization of restrictions on private distributors. 2.34 The current industrial policy environment has to be seen in the context of the controls dictated by the Central Plan over 1979 - 1986. The plan, coordinated by the Comissao Nacional de Plano (CNP) was applicable to the nationally registered enterprises, accounting for about 60Z of industrial employment. Targets were passed down from the CNP through the Government departments and the "Unidades de Direccao" to enterprise level. Annual plans, submitted by enterprises, went back up to the CNP and a process of material balancing took place so that by the end of the year the CNP provided finalized targets to enterprises. 2.35 By 1983 it was accepted by Government that a greater degree of flexibility had to be introduced into the system. Over the past two years the role of central planning has been diluted; the current situation is that the Unidades de Direccao are still expected to provide production data through Ministries to the CNP but the plan is now based on 'bottom up' submissions of production price and shipment figures rather than the 'top down' instructions of the CNP. Nevertheless, business behavior in the industrial sector is still inevitably constrained by the effects of controls - e.g. there is little procurement or marketing experience, nor experience of actively seeking opportunities for profit making. This applies to both the public and the private sector enterprises. E. Small Scale Enterprise: Structure and Performance 2.36 As stated above, the role of small-medium enterprises in the economy has only become a focus of Government concern relatively recently, and information on the sector has not been systematically compiled. The prospects for this sector are however of importance, and an attempt has been made to process some of the available data. The economy-wide scale structure of industry in 1973 and 1987 is shown in figures 2.2 and 2.3, and Anne._ 11. 2.37 The data for both of these figures contain omissions and underreporting. For example, small scale garment making is underreported in 1973, and activity outside Maputo is probably underrecorded in 1987. In addition Zambezia province was not surveyed in 1987. Bearing this in mind the scale structure still seems to have shifted decisively towards microenterprises (under 10 workers), with llZ of 1987 employment compared with 2Z of 1973 employment in this category. 1987 employment in microenterprises doubled compared to 1973. The total micro and small scale sector, defined as enterprises with below 100 workers, comprised 252 of total employment in 1973 and 31Z in 1987. At the other end of the scale, for - 22 - enterprises of over 500 employees, the 1973 share of 401 of employment has fallen to 29%, largely because of the semi-closure of industries such as cement, cashew processing and sugar. 842 of enterprises were in 1987 at the microlsmall scale level; extending the category to 'medium' scale (under 200 workers) would capture 95Z of enterprises with 46Z of total employment. By subsector the smallest average scale is in the wood products and textile (garment) industries (average 15 workers). The food sector also has a relatively large number of micro and small enterprises but also contains the largest (e.g. sugar), (see Annex 11). Figure 2.2 2.38 The overall 1973 estimate of 252 of employment in all recorded enterprises of INOUSTRIAL ENTERPRISE EMPLOYtEINT under 100 workers 9"Le Som 6, am." 59*9n a p p e a r s comparatively low; this could be 10-4B 2* accounted for by the known history Soo. /29.m Co.////of rural Mozambique (3955 S O* C9.mbefore independence where a combination of migrant labor and plantation \ / \ / agriculture may have distorted MD-4W C" M downwards the contribution of rural industry. The 312 estimate for 1987 however reflects the low c a p a c i t y utilization of larger enterprises as well aq a switch towards the smallest scale of production. 2.39 In the case of Maputo province alone (figures 2.4 and 2.5) employment increased between 1973 and 1987, from 43,000 to 46,400. Since the surveys of Maputo are likely to be more complete than those for other provinces, an analysis of the situation in Maputo province may shed more light on the issue. (See also Annex 11). 2.40 The figures confirm that in Maputo the only sector to exhibit significant growth, contrary to the overall trend, has been microenterprises with under 10 employees. There is also an increase in employment in the largest scale categories, which is principally explained by the establishment of textile spinning and weaving in Maputo province during the early 1970's. The structure of microenterprises in Maputo in 1987 was dominated by the garment industry, (consisting of 210 separate enterprises with 422 of total microenterprise employment). The next most - 23 - Figure 2.3 m p o r t a n t activities were engineering and metal products INDOJSTRIAL ENTERPRISE EMPLOYMENT (ssmall repair shops), and wood swA &AM ot -asc 7 products. 2.41 The small- medium enterprise sea. C29.2% 93 < { 10-49 CO. a s#ector, with or without the microenterprises, constitutes a significant subsector in terms of enterprises and 00\ 1t 9 C15.0 employment. 200-499 C24.Se However, the evidence of relative dynamism in the smallest f irms is not easy to interpret; given the quantitative restrictions on trade, and the Figre 2.4 shortage of foreign exchange and skilled manpower, it is possible that the micro- I NDUSTR iAL ENTERPR I SE EMPLOYIMENT-MAPUTO enterprise sector, >so" KRSammW wn w i t h 1 o w e r dependence on 10-9 C~W)imports and technical skills, SeW (2) 4 ///// 8gained market share even with products so-s l4090.m t h a t w e r e uneconomic (with high relative economic cost in relation to q u a 1 i t y ) . 2 \ Nevertheless, the apparent fact of solid growth in this sector is consistent with international experience and it is likely that micro/small scale - 24 - iagure 2.5 enterprises are . ~~~~~~~~~~g e n e r a 1 1 y potentially economic as a result of their INOLSTRIAL ENTERPRISE EYE-MAUT ability to utilize capacity more CS.%IW ~~fully, and their more appropriate lIIIj' g¢> ff^9X5pattern of ractor utilization, - C" SX3 | including use of s l / >"s0. * indigenous skills and managerial knowhow. In terms o f t h e reestablishment of longe r - term investment lending in Mozambique, there is therefore probably some justification for identifying a .__ _ _ _ _ _ _ __ _ _ _ _ _ _ _ __ _ _ _ _ _ _ target group of more managerially and technically progressive enterprises within these subsectors which are looking to 'graduate" - i.e. to expand capacity from workshop-scale. However, the current predominance of larger enterprises in terms of productive capacity means that the latter will continue to contribute proportionately greatly to the growth of production. - 25 - II. THE CURRENT BUSINESS ENVIRONMENT A. The Business and Investment Climate General Background 3.01. Interviews with businessmen in Haputo, Beira and Nampula suggested that the Economic Rehabilitation Program (ERP), initiated in January 1987, is having a major effect on the conduct of business in terms of how prices are established, how goods are sold, how credit is granted and how business decisions are made. The role of the government, while still predominant, has shifted substantially from that of a detailed decision maker in the operation of industries to one more of oversight. The main function of the Central Government, as perceived by the enterprises interviewed, is now taxation and allocation of foreign exchange, and therefore must be conceived in terms of its ability to provide the right type of incentive framework. State-owned or intervened enterprises are increasingly being put on the same footing as their private sector counterparts. Since January 1987 they have been expected to rationalize their costs to 'ell at competitive prices in the market. The largely private sector firms interviewed for this study were, generally, enthusiastic about the changes pre-1987 to the present. One businessman stated: *we are relearning the gymnastics of the market". 3.02. Both private and public sector managers. entrepreneurs and officials are now able to critically discuss industrial policy. The Association of Private Sector Entrepreneurs (AEPRIHO) and the Industrial Association were founded in 1988. AEPRIMO has at present 270 members (mostly small producers and retailers). The larger private sector firms are more involved in the Industrial Association. The traditional Chambers of Commerce are not active. The Government has supported these new associations and expects that the new Industrial Association would play a major role in guiding Government decisions. Industrial conferences in 1987 and 1988 were held at which the Government support for an increased role for the private sector in the economy was spelt out. 3.03. A principal concern of the Government is to find ways of rationalizing its ownership in the industrial sector. The process has been started with the enactment of decrees on the ownership of intervened enterprises and the sale of production units. The shortage of local and foreign private capital will however require that Government participation remains significant in the short to medium term, contingent on the success of the ERP in improving investor confidence. 3.04. A further inescapable problem is the security situation which has effectively isolated the rural areas. This constraint affects industry both in terms of purchase of raw materials and sale of finished products. The perception of entrepreneurs interviewed was that the security situation has not improved significantly over the past four years. There was, however, a new optimism among those interviewed that the internal conflict would be resolved as the economic reforms remonetize the rural sector and reintegrate it into the national economy. Distorted pricing of agricultural products, nonavailability of manufactured goods, and breakdown of transport systems, have obliged rural producers to focus their activities on local barter trade - 26 - and sales on the parallel market, rather than supplying the towns through official marketing channels. Such a situation woula have assisted destabilization. 3.05. A number of indicators of the winvestment climate" may be identified: these include the trend in company startups and sales, investment proposals, consumption and output indicators. Such information is not systematically compiled and is inevitably patchy, particularly because the level of investment and new starts has been very low over the past few years. The following sections consider this indicative information. New Investment ankd Company Formation 3.06. Indications of recent upturn in activity may be seen in the renewed investor interest. Proposals to the Foreign Investment Promotion Bureau (GPIE) have increased steadily, as follows. Table 3.1: MOZAMBIQUE - FOREIGN INVESTMENT PROPOSALS APPROVED BY GPIE (As of November 30th, 1989) 1985-86 1987 1988 1989 Total est est 10 5 20 21 56 3.07. The subsectoral breakdown of approvals was as follows. Table 3.2: MOZAMBIQUE - FOREIGN INVESTMENT APPROVALS BY SUBSECTOR (As of November 30th, 1949) Number of Total Foreign projects Investment Equity (US$million)(US$million) Agriculture 11 74.8 11.8 Tourism 11 42.6 4.5 Petroleum 2 20.0 20.0 Transport 14 12.3 3.7 Mining 6 2.2 2.2 Fishing 4 1.7 0.9 Manufacturing, cmrc. 4 3.2 1.1 Other 4 0.7 0.4 Total 56 157.5 44.5 Source: GPIE - 27 - 3.08. In the manufacturing/commerce sector the four approvals were for electronic equipment, refrigeration equipment, prefabricated building materials, and clothing. Approvals do not imply commitments, but the upward trend suggests that direct foreign investment inflows will increase. 3.09. A number of company sales and privatization at the national level have occurred over the past few years. In the light industry sector a total of 31 enterprises are reported to have been sold over the period 1983 - 88. These included principally wood products enterprises (11), plus oil and soap, leather, plastic products, agro-and food-processing, salt, beverages, metalworks, and textiles and garments; 6 enterprises in the last category were sold, including 4 on a part-ownership basis. Sell-offs in this category have increased over the period, with a bunching in 1987 and 1988. In the 'heavy industry" group a total of 16 company sales are recorded. These however include the electronic equipment subsector which consists largely of distribution and service agencies. Other enterprises sold included paper products and metal/engineering. The distribution of sales over time is not available, but there is clearly an accumulation of experience. 3.10. At the "local" level (i.e. for enterprises with approximately 50 workers or below) information is available on 40 company privatization of smaller industrial firms which were recorded over 1985 - 88. Of these 26 were in Maputo, 4 in Nampula, and 12 in Beira. largely in small wood working arnd metal working units. One 1988 sale was of a small chemical product and food processing company, priced at US$50,000 at the prevailing exchange rates. 3.11. There have been very few new company registrations at the national level in recent years and information is not available. New company registrations in Maputo at the 'local' level have been as follows; Table 3.3: HOZAMBIQUE - NEW ENTERPRISE REGISTRATIONS IN SMALL-SCALE MANUFACTURING PROCESSING AND REPAIRS - MAPUTO Subsector 1985 1986 1987 1988 1989 Total Food processing - 6 2 27 82 117 Garments 5 1 5 5 1 17 Leather - 2 - - 1 3 Wood, furniture 7 17 12 9 17 62 Soap, chemical - 3 - 1 - 4 Pottery, ceramics 4 1 1 2 8 Metals 4 17 7 6 6 40 Electrical 2 4 - 1 3 10 Total 18 54 27 50 113 262 Source: Provincial Directorate, MIE, Maputo. - 28 - 3.12. The majority of startups in 1988 and 1989 were cereal milling units for which a special subsidized credit line was available. New enterprises were largely in food processing (including milling), wood and metal work, and to a lesser extent garments. It is difficult to infer a trend, both because of the uncertain coverage of the statistics and because of the special conditions of the grain mill investments. However, the number of recorded new entrants over four years amounts to nearly 302 of the estimated total 1987 population of enterprises of under 50 workers in Maputo (see table 2.6). There are no data on closures for this group. Excluding grain mill investments 1986 was the peak year for new starts. A possible explanation is that this subsector had maintained a relatively high level of activity throughout the early 1980s, and is relatively unaffected by currunt changes. 3.13. There are general indications that both the small-scale domestic investor and the larger scale foreign investor are showing active interest in new ventures. Interviews suggested that some local investors were becoming interested in purchasing industrial assets at devalued prices. Enterprise Pricing 3.14. As stated, prior to 1987 all costs and prices of finished productu of nationally registered enterprises were controlled. The concern of managers -wa4 to fulfil centrally directed targets, although they were generally able to achieve only 20-30Z of plan because of shortage of raw materials or obsolescence of equipment. 11 Managers did not need to make decisions about marketing and their role was largely confined to carrying out day-to-day operations based on the allocations given by the central government. 3.15. Some of the private sector companies interviewed were able to cover their costs because the fixed prices of inputs compensated for artificially low selling prices for goods produced. This appears to have been the case for private sector firms in general (which in 1987 represented half of nationally registered firms). Losses were principally incurred by state companies. Companies generating losses were given automatic access to credit from the banking system. This led to insolvency as the debt and interest burden was excessive, both in terms of income and net worth. Private sector firms which generated losses operated in a very similar fashion to public sector enterprises. According to interviews with businessmen, private companies which produced essential goods (e.g. certain processed foods) were prepared to operate with negative margins in order to avoid the possibility of being intervened if production ceased. There was however a degree of confidence that, eventually, the Government would compensate these losses in an equitable fashion. 3.16. During the first half of 1987 a new price system was introduced with three categories: fixed, conditional, and free market prices. Initially, the new process for approving price increases was slow and cumbersome, both within the fixed price and conditional price category, but by the end of 1988 1/ Source: UNIDO, Industrial Situation in Mozambique May 1988. Vol.1 p120. - 29 - it had been reportedly speeded up. In the conditional price category the perception of firms is that they are free to set prices based on production cost with ex post reference to the relevant Ministry. The exception to this rule is a list of essential goods still under Government price control. 2/ One immediate result of the realignment of the exchange rate and decontrol of prices was that goods were redirected on to the official market. In those cases where industrial products are still price controlled official prices are also being regularly adjusted to reflect real costs. 3.17. In the case of raw material pricing, the impact of the 1987-88 devaluation had an immediate impact on the industrial sector due to its high dependence on imports. For the first time in several years the conduct of business required that expenditure on scarce imported materials had to be balanced against the abilil.y of the firm to generate sufficient working capital financing to fund the purchases, whereas previously, with overvalued exchange rates, a one or even two year supply of imported raw material would be easily financed, if available, at official exchange rates. Marketing and Distribution 3.18. Prior to the reforms introduced in January 1987 all industrial goods from nationally registered enterprises were supposed to be sold to government-designated buyers. The Government-granted right to purchase at artificially low prices allowed designated buyers to reap monopoly rents by selling these products (at & pre-determined price) to their clients who in turn had an incentive to resell (sometimes in collusion with manufacturers and sellers! or. the parallel market. Unlike agricultural products, which essentially disappeared from normal marketing channels (until the opening up the produce market in 1985), the industrial sector continued to produce and sell officially, largely because the industrial sector was more controlled by the Government, partly through its access to undervalued foreign exchange. 3.19. The all-pervasiveness of the parallel market prevented enforcement of official prices. In the case of tire production for example management estimated that in 1986 as much as 60Z of production was going to the parallel market. At the parallel exchange rate the dollar-equivalent price in Meticals was one thirtieth of world price, and the profit on diversion of sales across the border to Swaziland justified the risks involved. However, according to the firms interviewed, most industrial production was not smuggled cross border but sold within the country. For example, a food product manufacturer estimated that the products on sale officially at 16MT per kg. in 1986 (pasta) and 4OMT per kg. (biscuits) were obtainable on the parallel market at MT 200 per kg. and MT 750 per kg. respectively. In general, parallel market price for goods were three to ten times official levels. 3/ 2/ At the end of 1989, three groups of manufactured goods had fixed prices: gasoline and diesel, soap, and hydrocarbon derivatives (other than gasoline). In addition a number of essential products were controlled: flour, bread, sugar, cooking oil, beer, cigarettes. 3/ Source: IMF Recent Economic Developments November 1987. _ 30 - 3.20. For producers with low quality products the system of designated buyers was a major benefit as all production was guaranteed to be sold. This resulted in producers were not being held accountable for poor quality. Among manufacturers of consumer goods such as textiles and footwear the impact of decontrol was felt almost immediately with the accumulation of large Inventories of finished products once the designated buyers were no longer required to purchase stocks, because of inadequate quality at the new prices. These problems were aggravated by the imports during 1987 of similar, higher quality, low price products. This in turn forced local producers to rethink how the product was manufactured, how quality could be improved, how costs could be reduced and possible new outlets for sale. In many industries interviewed the creation of a marketing capability and the consequent search for new outlets of sale has occurred. Additionally, cost accounting was found to have become more important in the management of these firms While the quality of this information was questionable in some cases, overall it was clear that managers have a sharpened sense of the need to control costs. particularly in the area of raw materials. Capacity Utilization 3.21. With few exceptions, the industries interviewed continue to produce far below rated capacity. While capacity utilization has improved, some cases dramatically, from the very low levels in 1986, this provement is as yet limited. Most of the companies interviewed showed an erall increase compared to the low capacity utilization pre-1987. But as --terviews revealed, the quality, rather than the quantity, of products was becoming a main constraint. Perhaps the most significant change noted was that the constraint to capacity utilization was becoming a function of market demand and less of foreign exchange allocation. Interviews with enterprises yielded the following capacity utilization rates over times Table 3.4 4: 2OWBIQUE - CAPACITY UTILIZATION OP SELECTED COMPANIES Carmoc Metal Box Facobol Bicicl- Celmoque Berger Mabor (Boxes) (Cans) (Rubber) etas (Cable) (Paint) (Tires) 1985 15? 1oz 302 8Z 22? 13? 44Z 1986 19? 502 19Z 18? 162 272 1987 172 142 60Z 44? 42? 6Z 29? 1988 22Z 20? 60Z 66? 322 13? 482 Source: interviews. Current Industrial Output and Employment 3.22. As shown in chapter 2. industrial production started to expand in 1987, and has increased by over 25? during 1987-89. This was mainly due to _ 31 - the effects of donor import support programmes. 4/ The recovery was however spread unevenly between branches. Over 1985-88 there were major increases over a wide range, in beverages, textiles, wood products, printing, chemicals and plastics, nonmetallic mineral products, metal products machinery and equipment. However food products showed only a small increase (7S) over the period, and a number of products declined, cigarettes, clothing, leather, footwear, furniture, rubber products, and basic metals. The overall estimated increase in production in 1988 was 9.72. 5/ 3.23. The highly varied results stemmed from a number of factors. For food products (cereal milling) the lack of significant growth resulted from shortage of raw materials. In the case of cooking oil material shortages were combined with competition from donor imports . The overall increase in textile production arose from the partial rehabilitation of three mills (Texlom. Textafrica, and Riopele) over.1987-88 and donor-supported imports, but was partly offset by power shortages which reduced output of textiles elsewhere, and, in the case of blankets, donor-imports which undercut local prices. Garment production was another industry suffering from competition from cheap (second-hand) imports in 1988, but it recovered in 1989. PVC pipes and foam mattresses faced a 'aarket constraint in 1988. One difficulty noted was the erratic timing of import support, in for example metal products. 3.24. Comprehensive data for 1989 are not yet available, but preliminary estimates based on MIE projected results give a 7.6Z increase overall. Table 3.5: MOZAMBIQUE - OUTPUT GROWTN - K1M ENTERPRISES 1988-89 - change a/ Agroindustries 203.5 Rubber products 75.2 Food products 18.3 Paper products 2.6 Salt -4.0 Chemicals -19.8 Cooking oil 10.9 Glass -62.2 Drink 18.7 Construction machinery -39.1 Textiles 2.0 Electrical machinery -31.6 Clothing 21.8 Refrigeration equipment 14.1 Shoes, leather 11.8 Heavy engineering -6.1 Plastic products 9.4 Metal products 24.0 Furniture -4.9 Light engineering 6.6 Total (weighted) 7.6 a/ Calculated at 1985 prices. Sources KIM preliminary estimates. 4/ This included two Worid Bank Rehabilitation Credits which have allocated about US$60 million to industry. 5/ Source: CNP. - 32 - 3.25. In 1989 provisional data show significant improvements in food products, and in clothing, leather and rubber products, which had previously shown stagnation or decline. Textiles levelled out after rapid growth in 1987-88. It is notable that the major increases were in the consumer goods subsector, while intermediate goods recorded declines. From the pattern of growth and decline it would appear that, despite greater market orientation, output is still mainly determined by allocation and timing of foreign exchange inflows under donor aid. 3.26. The conflicting estimates of current employment were discussed in chapter 2. For example the 1987 labor survey showed employment of 66,500 in manufacturing, (with 17,700 in construction, and 3,700 in public utilities), covering all enterprises, while the National planning Commission estimates 87,800 in manufacturing for registered national and local enterprises, which would give an adjusted estimate of total employment in manufacturing and processing enterprise of about 110,000. It is not possible to assess with any accuracy therefore the effects of the ERP on changes in employment in industry. Discussions suggested that much of the overmanning of enterprises was eliminated by 1985. However, employment has not fallen as fast as output, and therefore it is likely that the some labor is currently surplus. Manufactured Exports 3.27. An important objective of economic adjustment is to increase domestic production and exports, particular'y from agriculture, but also from the other productive sectors. It was shown in chapter 2 that processed exports declined steeply with the decline in the major commodities. Manufactured goods exports have not been significant and have tended to be sporadic. In 1989, according to MIE, entports were running at an annual rate of about US$20 million, of which 40Z were petroleum product reexports, and the next catekjry were textiles and clothing. The weak export performance is despite the fact that some enterprises are now pricing their goods at below world price for broadly equivalent products. This was found to be the case for example for some food products and footwear. The principal problem, as noted, is one of quality rather than price; companies interviewed complained of the quality of packing materials, agricultural implements, bicycles, and inner tubes, for which import licenses were granted. The implication is that, in order to effectively reduce imports, local products need to be tradable, i.e. capable of substituting fully for traded goods. S. The Demand for Industrial Products 3.28. Industrial production continues to be influenced primarily by the allocation of foreign exchange for the purchase of raw materials, equipment and spare parts. However, in 1988 some industries appeared to facing market constraints which may have been continued into 1989 (although food processing enterprises are reported to have overcome such problems). The issue ef demand adequacy is important since there is no certainty that either the short or long term level of demand within the economy is sufficient to justify the retention of all existing capacity in some industries, in the absence of exports. Market difficulties for some domestic producers were caused over 1987-88 by apparently uncoordinated import licencing decisions, such as in _ 33 - the case of agricultural hoes, camestic metal utensils and garments. However more basic market concerns were expressed by some larger enterprises. Long-Term Demand 3.29. The level and pattern of demand in the long run may be considered to have been influenced principally by the effects on population and consumption of a) the natural growth of population; b) the settler exodus in the 1970's (possibly partly offset by the arrival of the foreign aid community); and c) internal and external migration caused by the security situation; indicative data are as follows: Table 3.63 MOZAMBIQUE - POPULATION AND CONSUMPTION (millions) 1973 1982 1987 Population (indig) 10.1 12.9 14.8 Settlers/expatriate 0.25 0.03 0.03 Refugees (external) - n.a 1.0 Displaced po,ulation n.a 1.7 "Affected' population 1.5 Urban population 0.7 1.3 2.7 Index of GDP 117.0 100.0 80.1 Index of consumption 112(est) 100.0 89.5 Sources$ IMF "Recent Economic Developm3ntsg 1987. 1988, and Staff Estimates. 3.30. The index of consumption shows a 10.5Z reduction over 5 years since 1982, and an estimated reduction of around 20? since i973. Settler departure may have reduced long term aggregate demand after 1975 by upto 20? given the considerably larger purchasing power per capita of the foreign community (concentrated on higher value manufactured goods). Refugee movement would have caused a reduction of about 8? in relation to the total potential 1987 demand, cutting the natural growth rate of settled population within the country over 1973 to 1987 from an expected 2.8?, to 2.12 per annum. Displacement would no doubt have been responsible for a further major reduction in consumption since those displaced would have lost their means of livelihood. The impact of the influx of the aid community would have partly offset the fall, but given 4ts high propensity to import this would may not have greatly affected internal market demand. 3.31. Abstracting from the effects of the dislocation and refugee problem, and assuming that the 200,000 departing settlers caused a reduction - 3 - of up to 20Z of the existing market for consumer goods, 6/ it may be estimated that aggregate consumption would have increased by 15 to 202 since 1973 given zero growth in consumption per capita. However, the market for higher value manufactures would not have risen as much, since the departure of the settlers would have affected this sector disproportionately. Thus it is possible that the market for some manufactured goods would have expanded little if at all over 1973 to 1987, even without the effects of the dislocation caused by the security situation. The security problem has created an additional depressant. Of the 40 enterprises interviewed most suffered in some way. Food processing and small scale industries located and dimensioned for urban markets like Maputo were less affected, but delivery of goods to and from rural areas is difficult and in some cases impossible. For example, one enterprise located in Nampula with capacity to supply cooking oil and soap to most of the Northern market of five million is unable, at present, to sell beyond the confines of the town, whose population is less than 100,000. 3.32. The implication of this rery preliminary analysis for the manufacturing sector is that potential demand for some industrial products will take some years to regain its 1973 level, even with the return of the refugees and reestablishment of peace. With the concinuation of security problems there appears to be no prospect of reestablishing 1973 domestic consumption levels for some consumer durable goods. However a more detailed study is required in order to establish the situation reliably. Impact of Structural Adjustment on Purchasing Power 3.33. The structural adjustment process also has important implications for industrial demand. Real devaluation would be expected to involve reduction of aggregate consumption, and expenditure switching via a redistribution of income and consumption from urban to rural areas. The effects of devaluation in lowering real wages and purchasing power would be expected to constrain the domestic consumer goods market, especially In urban areas that were not major recipients of aid funds. Rural demand for manufactures, which (in export commodity areas) wotld have been expected to have offset reduction in urban demand, have been to some extent out of reach because of disruption of distribution infrastructure. 3.34. The extent of expenditure reduction, and switching from urban areas would be a function of changing relative real incomes. Average monthly wage levels for urban workers reported by the industries surveyed for the study were MtlO,0O0 in 1986 and Mt27,500 in 1988. This category of wages therefore increased by approximately 1752. the increases are reflected in the index of consumer prices in table 3.7. 6/ This would imply that consumption per capita among the settler population was up to ten times that of the general population. - 35 - Table 3.7: MOZAHBIQUE - CONSUMER PRICE AND WAGE INDEX END YEAR 1986 1987 1988 Prices Food 100.0 220.7 299.9 Non food 100.0 269.8 442.2 Overall 100.0 263.3 395.1 Wages Average 100.0 180.0 275.0 Minimum (industrial) 100.0 220.0 500.0 Real vages (avg) 100.0 68.3 69.6 (mi"nmum) 100.0 83.6 126.6 Source: Informacao Estatistica. National Directorate of Statistics, and staff estlmates. 3.35. The overall consumer price index rose by 3952 from end-1986. just prior to devaluation. Food prices rose less than non-food prices in 1987, because the initial index included parallel market prices for food, whereas non-food (e.g. utilities) prices had been more effectively controlled and thus rose more steeply. Adjusting average wages for the rise in the overall price index would imply a reduction of real wages from end-1986 to end-1988 of around 30S. However, if minimum wages are taken as an accurate indicator then the level of real wages has actually risen over the period. The sharp decline in real wages in 1987 was reversed in 1988. This finding is however highly sensitive to the representativeness of price index weights and formal sector wage data. It is also sensitive to the timing of increases within the year. 3.36. Taken at face value rising real wage rates in the urban formal sector are somewhat unexpected; ceteris paribus it would imply that the loser from the real currency depreciation is the urban informal sector population, especially outside Maputo, and possibly parts of the rural sector producing controlled price commodities for the domestic market (such as maize and rice) who, unlike the producers of export commodities, would not gain from higher export prices 71. However, it is possible that the minimum wage and price 71 Ongoing research for the forthcoming Country Economic Report indicates relatively high gains for export commodities. - 36 - index are unrepresentative and a downward adjustment in the real wage estimate would result. 8/ To the extent that there has been an overall reduction in urban household real incomes, at least out.side Maputo, then discretionary purchases would be reduced relatively to essentials such as food. 9I One of the first expenditure categories to be reduced would be imported or import-intensive consumer durables such as electrical goods (e.g. radios), and bicycles; in addition manufactures such as textiles, clothing. cigarettes, and high value packaged foods would probably be affected. 3.37. Definitive information on market demand is scanty. Certain industries have faced a decline in production from 1986 levels which they consider to be due primarily to a retraction in demand. Principal examples were in cigarettes where one manufacturer stated that 1987 monthly production was 24 million cigarettes, down from 36 million in 1986. However this was partially reversed in 1988 because of relative price reduction. One footwear manufacturer reported that it was in 1988 producing only 501 of 1986 levels, more due to limited purchasing power than quality. Cardboard container orders declined to a significant extent after the July 1988 devaluation and price increases. Retraction in demand was also noted in the case of tires, oils, soaps and bicycles, and PVC pipes. Foam mattress demand fell sharply as consumers switched to lower quality substitutes. Large stocks of radios built up over 1988, of which 702 were exported to Bulgaria. Of the enterprises interviewed in Maputo in 1988, those mentioning demand as a present or near future concern were mainly higher value consumer and durable goods. 81 The price index could be distorted by unpredictable supply of controlled price goods and rapid increases in costs of energy. The rise in the minimum wage index may be overestimated if informal sector incomes have lagged behind increases in minimum wages. In the case of the lowest income groups, accounting for up to 50Z of urban households, about 902 of expenditure in early 1989 went to food, energy and transport, with the remaining 101 available for consumer goods (apparel and household supplies). See R. H. Green Social dimensions of adjustment; poverty assessment paper Government of Mozambique, February 1989. 91 This would happen because of the relatively high income elasticity of demand for discretionary goods as compared with essentials such as basic food, and energy. - 37 - Table 3. S MOZAMBIQUE - ENTERPRISES REPORTING DENUD CONSTRAINTS Name * Location Product Employment Mabor Haputo P Tires 550 Hadal Haputo P Margarine 3,500 Cia. de Bicicletas Haputo P Bicycles 180 Riopele Maputo P Synthetic Textiles 800 Facobol Maputo P Rubber products 390 Grupo Entreposto Maputo P Cooking Oil, Soap, Caju 3,000 Protal Haputo P Condeused milk 180 Textafrica Haputo P Textiles 4,000 UFA Haputo I Rubber footwear 480 SAT Haputo P Cigarettes 140 Emb. Holdains Haputo P Cardboard containers 66 Carmoc Maputo M Cardboard containers 250 * P (private); M (mixed); I (intervened). 3.38. Finished goods stock levels provide some indicator of the balance of demand and supply. Available data appear to show that stock levels built up to a peak at the end of 1987 and during 1988, concurrently with the decline in real wages, and generally declined in 1989 as wages caught up with prices. Table 3.9: MOZAMBIQUE - FINISHED GOODS STOCKS; SELECTED INDUSTRIES Product Canvas Cigarettes Domestic Batteries Radios shoes Utensils Unit 000 prs million 0008 units units 1987 137.7 98.5 98.8 5242.8 37441.3 1988 161.7 89.4 235.2 6488.5 49530.7 1989 83.3 65.1 215.9 5775.7 24989.3 3.39. 1989 average quarterly stock levels were in all cases below those of 1988, and in the case of shoes and cigarettes below those of 1987. For utensils, batteries and radios destocking is not as clearcut, and radio stocks, as mentioned, were run down as a result of a single export order. Given averaged production levels, this tends to support the argument that durable goods in particular are demand constrained. _ 38 - The Impact of Donor Aid on Production and Markets 3.40. Donor assistance to industry has been running at approximately US$ 100 million per annum over 1987 - 89. In some cases assistance has gone to rehabilitate plant and equipment, but the majority of donor assistance has provided foreign exchange for recurrent imports of industrial materials and spare parts. The impact of donor support has had short term benefits in terms of permitting existing enterprises to increase capacity utilization. However, there are adverse impacts if donor goods undercut the prices of import competing local enterprises burdened by short run operating inefficiencies. The donor import support program also involves problems for recipient enterprises in terms of tying of procurement and unpredictability of supply. Procurement tying potentially reduces the value of donor assistance to the extent that tied supplies are above international price in relation to quality or specification. 10 The problem of unpredictability of supply of donor inputs is also important, since effective management requires planning ahead. Donor supplies and emergency procurement from abroad under assistance programs have reportedly caused marketing difficulties in the production of blankets, soap, and also agricultural tools, while in 1988 donations of second hand garments caused similar problems. Thus poorly planned commodity aid may have adverse effects on both production and the market for industrial goods, on the one hand because it may lead to increased domestic costs and, on the other, because it may undercut domestic prices. C. The Reform of Industrial Policy Reforms Under the Economic Rehabilitation Program 3.41. This section summarizes the most important structural reforms (as opposed to short-term stabilization measures) affecting the industrial sector that have taken place or are in process under the Economic Rehabilitation Program. Some of the changes are addressed further in the following chapters. The reforms may be conceptualized as (a) those relating to the macroeconomy, and (b) those which are more sectoral or enterprise-specific. 3.42. Exchange Rate and Foreign Exchange Allocation. A number of measures have been taken by the Government. An export exchange retention scheme was introduced in 1983 which entitles exporters to hold a variable proportion of their proceeds at the Bank of Mozambique in a convertible account, with an average currently of about 502. As stated, currency devaluation appears to have lowered the real value of domestic factor payments such as wages, and improved the medium to long-term competitive position of labor-intensive and domestic resource-based industry in relation to imports. At the same time markets for some manufactures have been depressed. Currency devaluation should allow a phased opening-up of the 10/ For example, if each USS million of goods or services carries a 'donor price index" 25? above the lowest sustainable world price of the equivalent goods or services, then the effective value of goods/services supplied is $1,000,000/1.25. or US$0.8 million. _ 39 - foreign exchange market; this process has been supported by a System for Non Administered Allocation of Foreign Exchange (SNAAD). 11/ The Government has also created two administered funds - the Market Fund for the trade sector, and the Small Enterprise Fund, providing foreign exchange to industrial enterprises. As the value of domestic currency falls to more realistic levels the coverage of the SNAAD is expected to widen, and special funds may be phased out. However, a significant proportion of foreign exchange is still allocated from aid sources. (Implications of the system for the allocation of foreign exchange are discussed further in chapter 5). 3.43. Pricing, Distribution and Profits. Decontrol of the majority of product prices has assisted in eliminating enterprise operating losses which were attributable to fixed prices up to 1986 and which were financed for some years through bank lending. Three categories of price are allowed - fixed prices, subject to regular review, conditional prices based on a cost-plus formula, and free market prices. Since January 1987 policy has aimed at a) reducing the number of products with fixed prices and b) allowing prices to be more related to costs of production and market demand. Decontrol of purchasing and distribution by enterprises, which was previously subject to central planning directives, has resulted in an expansion of marketing and procurement activity by enterprises, and a greater awareness of the costs of financing stock holding. Distribution policy is expected to continue to change alongside pricing. The number of final products subject to administrative allocation was eleven by December 1989, and is expected to continue to decline. All enterprises are now permitted to trade directly rather than through a trading company; state trading companies are expected to compete with the private sector. However, control on foreign procurement is still exercised through foreign exchange allocation and import licencing, with control of some internal trade through rationing. 3.44. Tariff and Tax Reform. These reforms have begun to rationalize the system of taxes facing companies, to simplify the trade tariff system and subst'-.. te limited tariff protection for quantitative controls. As a result of discussions with the business community the Government made changes in the corporate tax system in 1987 and 1988. This included the abolition of dividend tax. During 1989 and 1990 simplification and rationalization of customs tariffs is also under way. There remains a relatively complex set of taxes on business, including Industrial Contribution (502 of taxable profits); Complementary Tax (5X to 702 of global income); Turnover Tax (5Z on enterprise revenue). Under the foreign investment law of 1984 and the domestic investment law of 1987 there is also a system of exemptions, including tax holidays, depreciation and capital allowances, and exemptions from labor income tax, turnover tax, and import and export duties applicable to investment projects. (This is discussed further in chapters 5 and 7). 3.45. Investment Incentives. Efforts to attract foreign and domestic private investors have been being made, through the establishment of trade associations and through privatization and joint ventures. In 1984 and 1987 11/ This provides a "window' for imports of industrial inputs initially for the textile and footwear industries, and spare parts for construction equipment, agriculture and transport. _ 40 - the foreign and local investment codes relating to incentives and foreign exchange retention were introduced. In December 1988 a debt-equity swap law was passed setting out conditions under which foreign investors could acquire local assets. However, the actual impact of the incentives is difficult to judge in current conditions where investment is constrained by numerous other factors. Special arrangements have been negotiated by some foreign companies, and a case-by-case approach to the incentives on offer is generally practiced. The system of corporate incentives is further discussed in chapters 5, 6 and 7). 3.46. Enterprise Financial Restructuring. A number of enterprises require financial restructuring, firstly because of the effects of devaluation on balance sheets, and secondly because of the extent of outstanding debt which many formal sector enterprises accumulated up to the end of 1986 as a result of price fixing, as well as other factors. By December 1988 the Government had taken steps to transfer about Mts 8.0 billion of the debt of 10 state and intervened companies to the Central Budget, where revised terms and conditions of repayment are to be worked out. 12, Transfers up to now have been confined to public sector companies, and private debt is as yet not being transferred. Some individual debt reschedulings have also been arranged through the Bank of Mozambique, on easier terms. A decree has also been issued requiring revaluation of assets of enterprises, in order to assist in reestablishing the basis for industrial financing. (See chapter 6). 3.47. Accounting and Auditing Regulations. High quality accounting and auditing is critical to the efficient operation of business. The regularization of the banking system and the regulation of the private sector require auditing capacity to be set up. A system of standard accounts was instituted in 1984, but there is an acute shortage of accorntants and almost no independent accountancy profession. Currently, external auditing of enterprise accounts is not required; internal auditing is carried out by some firms and auditing for tax purposes is undertaken by the Ministry of Finance. However, the reform process has encouraged foreign accountancy firms to set up offices. A local firm has started to carry out company valuations on behalf of the GPIE and a number of local firms with some accountancy skills have started business since 1987, largely aimed at the foreign aid market. (See chapter 7). 3.48. Company Sales and Ownership. The original legislation relating to the intervened companies has been supplemented by decrees of 1988 and 1989'. 13/ The latter decree sets out the procedure for sale of companies designated by the authorities, involving valuation and processing of bids for purchase. National and provincial Appraisal and Sales Committees consisting of representatives of the Government, banks, labor union and licencing body 12/ Potentially over Mt4O.O bn may be transferred, to be finafced from the counterpart funds of the IDA Rehabilitation Credits. 131 These are decree 145/88 of November 1988 and 21189 of May 1989. Among the provisions of the decree are the stipulation that state ownership must be formalized prior to sale (Article 5). - 41 - have been established to implement company sales. Up to now the Government disposed of assets abandoned by previous owners under 1977 legislation, and has coAe to agreements with new owners on a case-by-case basis. For industrial enterprise sales, asset valuations and the structuring of sales agreements were carried out by working groups convened by the Ministry of Industry. Sales decisions were guided by two main criteria; th.rt the enterprises maintain the activity carried out prior to the sale (failing which a new licence is required), and that employment would be maintained within the requirements of the labor law. The 1989 law stipulates the first of these requirements. An analysis of some mechanisms for company transfer and privatization is set out in chapter 7). 3.49. Labor Legislation and Wages. Until January 1987 there were 4 basic occupational categories. This system was inflexible and detracted from efficient use of labor. In 1985 a considerable amount of labor shedding took place and in January 1987 a new labor law was enacted. 14/ This set twenty grade ranges in industry, specifying new pay levels and suoplementary payments such as performance-linked and seniority bonuses. The aim is to permit more flexibility in employment, in terms of rewards for efficiency and ability of enterprises to lay off workers. It is intended that these measures will allow wage costs to more accurately. reflect operating conditions. Hiring and firing is now acceptable subject to referrals to the Ministry of Labor. However, it appears that labor laws remain relatively restrictive from the point of view of investors, by international standards. 151 During the period when the law was being implemented, a series of interim wage increases were permitted which offset the effects of price increases. General industrial costs have probably fallen in real terms due to devaluation; however, where overmanning still exists the longer run labor cost levels of enterprise would still be high. 161 14/ The four-level employment classificat:.on was governed by Law 4/80 of 1980. The new law was enacted under decree 5187 in January 1987. 151 Under the current legislation workers receive a lengthy warning period of redundancy, amounting to several months, with longer periods depending on length of service. For international comparisons see "Report on the feasibility of export processing zones in Mozambique%. World Bank April 1989. 16 Current average industrial wages are in the range US$18.0 to US$36.0, at, respectively, the parallel and official exchange rates. The lower level, which is probably the better indicator of relative purchasing power, is now low by the standards of the poorer competing third world countries. The 'Report on the feasibility of export processing zones in Mozambique", IDA 1989: op cit, cites wage levels of US$25 to 30 per month in the Philippines and China. Real wage levels for industrial labor are also below those of most Southern African countries. The most critical indicator is however the real wage in relation to labor efficiency. _ 42 - 3.50. Competition Policy; Industry Entry and Exit. The entry and exit conditions for industrial enterprises depend inter alia on the ability to acquire and transfer real and financial assets, and the mobility of factors of production. The regulations regarding company purchase and sales, and laws regarding hiring and redundancy of labor are critical aspects of this. The experience in company startups has been verv limited in recent years, except in the small scale sector, and procedures are unclear. Experience with company sales is greater however, and the new laws regarding disposal of state-owned companies have supplemented the 1970's legislation on interventions and abandoned companies. There is also a Commissao de Reorganizacao das Empresas 171 which is considering these issues. Under the law of 1989 there now appears to be no formal constraint on sales to the private sector and the Hinistry of Industry is establishing a bureau for divestment of interviewed enterprises. The number of privatization which have been permitted, and the interest in joint venture mechanisms, suggest that entrylexit conditio.s have improved substantially. This seems also the case wlth the banking system, given the increased 1988 share of the private bank and the development cf a securities market, discussed in chapter 6. However, the revised employment legislLton may remain unduly restrictive. The laws regarding landholding, which are not discussed in this report, are also i-portant to factor mobility and raising of investment finance, and should be the subject of further study. Reforms and Medium-Term Strategy for Industry 3.51. Despite the series of measures introduced, a supply response from -~' ry cannot be expected in the short run (apart from increased ation of existing capacity due directly to increased supply of inputs lonor import programs) because of the rundown state of plant and aent, lack of skills and managerial resources, and the breakdown of ing and distribution systems through sabotage and disuse. For existing rrises a two phase strategy is called for, involving initially short Lc." rehabilitation, and later, after the economics of the operating industrial sector become clearer, major refurbishment for economically viable producers. 3.52. A difficult transition is under way from a supply-oriented planned industrial sector to a market-oriented environment in which production and investment respond to market and price signals. This involves setting up the enabling conditions for efficient investment rather the specification of a particular pattern of production or investment. Underpinning this transition the process of price (and wage) adjustment must continue in order to allow costs and prices to vary according to domestic market conditions, and secondly in order to permit domestic pri.ces to converge over time with international prices, so that an efficient distribution of production between domestic and international markets is attained. This requires continued action in main policy areas of the ERP, namely: currency realignment, and foreign exchange access, rationalization of corporate taxation and trade 17/ This commission is chaired by the Minister of Industry and operates with the representation of the Minister of Justice. _ 43 - tariffs, financial reorganization (of enterprises and banks), and clarification of the corporate regulatory system. 3.53. In chapter 2 it was shown that the micro/small scale industry sector has been particularly adaptable to adverse conditions and its role is likely to be of special significance. Since 1983 industrial policy has emphasized more efficient use of existing capacity in larger enterprises, and a shift of emphasis to small enterprises, as is evidenced by establishment of Government support agencies, the special foreign exchange funds. Although there is no systematic bias against smaller enterprises, as has occurred in other countries, in areas such as foreign exchange allocation, fiscal incentives, and access to credit, smaller enterprises do however lack access to information about incentives, and their less formal setup, often outside the tax system, may exclude them from the priority lists drawn up under administered schemes such as for foreign investment allocation. In general the price reform process should favor domestic resource using, labor intensive production, which is characteristic of smaller enterprise activity. ; Smaller enterprises using local resources should be able to mnintain a j relatively high level cf capacity utilization in the continuing situation of foreign exchange scarcity, provided that they not face, rehabilitive administrative constraints. 3.54. *dual technology' approach is called for, developing the existing capacity of efficient larger enterprises, which make the major contribution to current output, and taking advantage of the favorable factor utilization and learning potential of the smaller enterprises. The incentive system should support this strategy. - 44 - IV. EST3MATION OF THE ECONOMIC EFFICIENCY OF INDUSTRY A. Industrial Efficier-cy Analysis 4.01. Prior to the implementation of the Economic Recovery Program, the cost structure of Mozambican industry in international terms was distorted by the overvalued exchange rate and its effect on internal relative prices. Devaluation appears to have reduced domestic costs to more competitive levels and therefore improved the economic efficiency of industry. However industry would still appear to be operating at low levels of technical efficiency because of low capacity utilization, generally obsolescent plant and equipment, and in some cases inefficient use of labor and materials. 4.02. Despite the current low productivity, few of the currently operating enterprises seem to have been of fundamentally uneconomic scale or technology at the time of establishment, which was carried out largely by private local and foreign investors up to 1974, when profitability considerations were the primary investment criterion. Most plants are therefore likely to be of a scale appropriate to a peacetime domestic market. Many produce goods whose high transport cost or perishability provide them with natural protection against imports. 4.03. Many enterprises are operating with fully depreciated plant and equipment of which a surprising amount continues to function. To be financially profitable in the short run these enterprises need only to cover direct costs of production (i.e. excluding depreciation cost). Given these circumstances, relatively small investments in spare parts and upgraded equipment could probably produce high marginal returns in many enterprises over the short run. Over the longer term, much higher levels of reinvestment in modernized plant and equipment will be required. For existing enterprises rehabilitation must therefore be viewed as an essentially two-phase process: (a) opportunistic, cost-effective partial rehabilitation of selected enterprises over the short run; and, (b) more complete capital and human resource investment to sustain modernization and expansion over the medium and long term. 4.04. In order to assist in prioritizing subsectors and enterprises for aiter alia foreign exchange allocation and investment promotion, it is necessary to have a method of assessing efficiency of resource use. A simple measure of efficiency which seems appropriate is the domestic resource cost (DRC) ratio. This section reports on and interprets the DRC ratio results obtained from analysis of the accounts of a sample of enterprises; technical appendix 4.1 illustrates the methodology in more detail. Sector Ministries have made DRC ratio calculations in the past, although they have not used for investment decisions. The ratio is defined as: Domestic (nontradeable) resource cost Net foreign exchange effect - 45 - 4.05. DRC ratios measure the cost of domestic resources used in the production of a product per unit of foreign exchange saved or earned. 1/ Firms use both domestic and imported -esources in the production of any good and in the process they either save foreign exchange through import substitution, or earn foreign exchange through exports. Most industrial firms in Mozambique substitute imports and hence may save foreign exchange. The DRC measure focuses on trade efficiency and implies that foreign exchange is the critical constraint on development. 2/ For the foreseeable future this is likely to be the case in Mozambique; thus the use of an investment criterion which maximizes net foreign exchange earnings and savings per unit of domestic resource would be appropriate. 4.06. DRC analysis can be used to measure either short-run efficiency (current operations) or longer run efficiency of a subsector/enterprise. The short run DRC ratio is based on direct costs alone while the long-run DRC also takes account of plant depreciation and replacement. The short-run ratio would be relevant in the allocation of investment for recurrent inputs and minor rehebilitation of existing plant and equipment, and the long-run ratio would be appropriate for considering the allocation of funds for expansion or more major plant replacement. 3/ 4.07. If a subsector or enterprise is found to be economically inefficient, it is important to distinguish whether the source of the inefficiency is exogenous, resulting for example from the effects of the exchange rate/tariff system, or whether it is due to technical problems within the subsector/enterprise (e.g. in processes, labor, management, material use). If technical inefficiencies are the basic problem then technical reorganization (if feasible) would be required in order to increase efficiency. If technical productivity is satisfactory (that is, if enterprises are operating near full capacity with efficient use of inputs) then changes to the overall industrial policy framework (e.g. exchange devaluation) may be the solution - i.e. they may lower costs to internationally competitive levels. 1/ If an enterprise uses more than one unit of domestic (nontradeable) inputs to save or earn a unit of foreign exchange then its DRC ratio would be above 1.0, indicating that it may not be efficient. A firm that uses less than one unit of domestic resources per unit of foreign exchange earned or saved would have a DRC ratio of below 1.0 and would be efficient. (See also note 10 of this chapter, and technical appendix 4.1). 2/ As such it is complementary to other measures currently used by planning departments in the Government: e.g. the capital/employment, capital/value-added, and domestic value- added ratio. The latter measures could be used in conjunction with an economic efficiency measure. Ideally it is also desirable to calculate both ratios at a) existing capacity utilization, and b) at maximum feasible capacity utilization. _ 46 - The Enterprise Survey Industrial Efficiency Estimates 4.08. As shown in chapter two, the industrial sector consists of approximately 875 enterprises with ten or more employees, with the heaviest concentration of industries located in Maputo, followed by Beira. Agro- industrial and forestry operations are, obviously, more dispersed. For the estimation of industrial efficiency a sample survey of 62 enterprises was carried out, in conjunction with interviews relating to the qualitative aspects of the business environment. 45 enterprises were included in the analysis; those which were excluded (many of the smaller firms) lacked appropriate cost information or made products without identifiable world price comparators (e.g. household furniture). 4/ Estimation was sensitive to assumptions about prices and product comparability. The rapid price changes over 1987-88 meant that care had to be taken to ensure that the cost and output value data referred to equivalent dates. The subsector distribution of the 45 enterprises analyzed was: SIC 31 - Food, drink, tobacco - 8 SIC 35 - Chemicals & plastics - 5 SIC 32 - Textiles, footwear - 11 SIC 36 - Non-metallic minerals - 2 SIC 33 - Wood products - 2 SIC 38 - Metal productsl SIC 34 - Paper and paper prod - 3 equipment - 14 4.09. Original data used for the analysis consisted in most cases of enterprise production cost estimates at average capacity utilization, which had been used to make a case for proposed price increases. Official exchange rates at the time the data was prepared were used for evaluating costs of traded inputs. Where lack of price data or comparability problems occurred, supplementary information was used to provide approximate quantitative estimates. 4.10. Tables 4.2 and 4.3 summarize the results of the DRC analysis for the 45 enterprise sample. In each table coluum 1 lists the estimated DRC ratios by product. Coluum 2 lists either actual prices or, where price information was inadequate (in table 4.3), 51 the hypothetical CIF price of a comparator at a DRC ratio equal to 1.0. If the domestic cost is lower than the hypothetical price then the enterprise would be efficient. For textiles, garments, and metal fabrication, inferences of subsector efficiencies were made using alternative information on comparative prices, export potential, rehabilitation potential, etc. The approximate capacity utilization rates 41 The survey was carried out in two phases during 1988, with a total of 62 enterprises visited, with representation from each major subsector. The sample design should ideally have been stratified on the basis of either value added or employment distribution, and scale considerations, but there was in the end a bias in sample composition towards larger firms with more complete accounting records. Due to the quantitative restrictions on most imports which compete with local production, price estimation for comparators was more difficult than it would be under a tariff-based protection system, or a liberalized trade regime. - 47 - prevailing are shown in Column 3. 6/ Columns 4, 5 and 6 indicate respectively for each product the import content as a percentage of total factor cost of production; labor as a percentage of total production cost, and labor as a percentage of domestic resources used in production. 4.11. Tables 4.1 and 4.4 summarize the results of a separate study of ten entergrises where more detailed cost and price information was available, including estimates of rehabilitation investment costs. 7/ Consequently it was possible in this case to divide the results into (a) short term DRC ratios (table 4.4), and (b) medium term ratios (table 4.1). 8/ 4.12. The enterprises listed in table 4.2 and table 4.4, for which relatively firm price information was available, mostly showed short-run DRC ratios of less than 1.0. This category is therefore efficient and short-term investment in inventories, spare parts, or minor plant refurbishment would be justified. In addition, if the official exchange rate is overvalued, a DRC ratio of above 1.0, reflecting a "realistic* exchange rate, would also indicate an economically efficient producer. 9/ In this category seem to be timber processing, canvas shoes, and copra oil/soap, which would therefore also be economically efficient. Apparently inefficient enterprises included sugar, bicycle inner tubes, and metal pipes. Cement production frcm iuported clinker is also uneconomic (because of transport costs). Production of inner tubes at the lower assumed price ($0.78/unit) shows a negative DRC ratio. This would imply a negative foreign exchange effect; i.e. that the country 6/ In principle capacity utilization rates should have littli effect on short run DRC ratios which only taken into account direct costs. However, an element of fixed costs would remain; e.g. overhead (fixed) labor cost per unit of output would vary inversely with capacity utilization. 7/ This review was the work of Arthur. D. Little, consultants to IDA for the Industrial Enterprise Restructuring Project. The ratio is medium-term rather than long-term because estimated investment and production costs and revenue were for rehabilitation rather than complete replacement of equipment. 91 If the realistic exchange rate is 30Z above the official rate, then a DRC ratio of up to 1.0 + 302 (1.?) is acceptable. Illustration: an enterprise incurs costs of Mtl billion per annum for domestic (nontradeable) inputs, and US$1 million for imports, and its production has a World price value of US$2 million. It therefore uses domestic resources of Mtl billion to save foreign exchange of US$l million, yielding an 'own exchange rate" of MtlOOO to US$l. At an official exchange rate of Mt9OO to US$1, MtlOOO of costs would be excessive and would signify "inefficiency". However, if the realistic rate is 302 higher (Mt170 per $) then production would be efficient. To interpret the DRC ratio it is therefore necessary to take into account the exchange rate at which the ratio has been calculated. - 48 - was spending more foreign exchange by producing tubes than by importing them directly. In all the above cases capital costs are treated as sunk and are ignored; when capital costs are taken into account the results would tend to differ. 4.13. The supplementary analysis (table 4.3) shows that other subsectors, (such as textiles and glass) may not be cost competitive in the present circumstances. However, others such as garment production based upon imported fabric, metal working and metal fabrication are probably efficient, due to their relatively high labor-intensity. The high DRC ratios for sugar production result from security problems, disorganization of plantations, and obsolete equipment. 10/ (See also technical appendix 4.2.) 4.14. Table 4.1 shows DRC ratios calculated taking into account rehabilitation investment. All the ten enterprises surveyed, some of which are multiproduct, are economically efficient. Table 4.12 MOZAWBIQUE - DOMESTIC RESOURCE COST RATIOS (MEDIUM TERM) Ten Enterprise Survey DRC ratio Agricultural implements 0.79 Metal pipes and tubes 0.21 Oils and soaps 0.93 Shoes, tires and tubes 0.96 Cartons and board 0.70 Cement 0.92 Steel drums 0.50 Foods, animal feeds 1.05 Copra, margarine, feeds 0.86 Bricks, stone 0.48 Sourcet estimated from report br Arthur D. Little for IC. 4.15. The results of table 4.1 are based on feasibility study projections and assume certain targets for capacity utilization and productivity; they should therefore be interpreted as possibly achievable, not actual, levels of efficiency as distinct from levels of efficiency based on existing capacity utilitization. The results in general are superior to the short run DRC r--sures. Under the assumptions used metal pipes are economically efficient, whereas in the short run, on the basis of current operating conditions, they are clearly not (see table 4.2). Taken at face 10/ A sugar industry subsector study is planned under the IDA Industrial Enterprise Restructuring Project. Bilateral donors have also shown interest in pre-investment rehabilitation studies for individual mills. - 49 - value these results imply that rehabilitation investment can be justified for all the enterprises studied; this conclusion is reinforced if a lower US$- per-Mt exchange rate is assumed. (See note 10). However, there is likely to be an optimistic bias from prefeasibility studies and this would raise these ratios. 4.16. The possibilities of successful rehabilitation is strengthened by the fact that industrial labor appears now to be relatively low-cost in Mozambique. Tables 4.2, 4.3, and 4.4 show very low ratios of labor cost to total costs. The prevailing average wage for factory workers in 1990 is in the range US$18 at the parallel and $36 per month at the official exchange rate, plus benefits. This compares to international rates of US$25 to 30 in the poorer countries of A&ia, (Philippines, China) which would be the main competitors in for example garment exports. 11/ 4.17. The relative efficiency of the surveyed industries does not appear to conform to any identifiable pattern; for example it cannot be stated prima facie from tables 4.2 and 4.4 that inefficient industries are from relatively highly import-intensive or capital-intensive subsectors (as can be seen from the labor cost items in the tables). The most likely explanation for this is simply that in the current very difficult operating conditions expected comparative advantages are not reflected in short run costs. However, the analysis does suggest that the plants which give indications of being inefficient tend to be flow or continuous-process operations and where there are economies of scale. These would be the first to be affected by the frequent shutdowns, and other operational problems of recent years. Conversely, the more efficient operations tend to be those which use batch production and piece work. 4.18. The measures of efficiency in the current conditions of Mozambique industry (with low capacity utilization and severe technical problems) are not sufficiently reliable to lead to clear indication of investment priortization short run DRC results are consistent with recent estimates for Zimbabwe, Tanzania and Malawi in that they demonstrate short-run efficiency. However, in the case of the long-run estimates, (table 4.1) the levels of efficiency are comparatively high. This is likely to be because the estimates are based on the assumption of prior rehabilitation and higher levels of capacity utilization. D. Conclusions 4.19. In conclusion, the results of this preliminary analysis suggest that: (a) Mozambique''s industrial sector cannot be dismissed as "inefficient'. It appears that both short-term investment in recurrent inputs and longer term rehabilitation is feasible for 11/ Sources Report on the feasibility of an export processing zone in Mozambique World Bank 1989. In fact the relevant comparison should be in terms of labor cost per unit of output. - 50 - would include low capital cost, labor-intensive fabrication and assembly operations in the Maputo area, and a range of industries which have a degree of natural protection, such as building materials. (A fuller discussion is in technical appendix 4.2). (b) It is Important to consider industrial rehabilitation priorities through a case-by-case study of individual firms. It is only through detailed enterprise level work that the root causes of the industrial inefficiency can be identified, and appropriate rehabilitation designed. (c) Efficiency analysis is not very instructive in the present industrial. environment in Mozambique. However, when a greater degree of economic stability is achieved, and enterprises are operating near to their capacity limits, DRC ratios could provide indicators of the direction that long term investment should take, and a more rational system of protection. In the short to medium term such efficiency measures could provide an economically rational system for foreign exchange allocation for imported inputs to enterprises. These issues are discussed further in chapter 5. 4.20. A regular system of efficiency measurement could be introduced if a database of world prices (imports and exports) was accessible. Several Govqrnment departments already collect such information and it would be possible to systematize it. This would be particularly useful after achievement of a greater degree of price and production stability, and after the expected introduction of trade tariffs to substitute for quantitative restrictions, when more realistic assessments of production costs and competitiveness would be possible. 4.21. Assessment of economic efficiency assumes some capacity- utilization norm. Prom chapter 3 it can be seen that expenditure reduction has tended to depress demand for manufacturing and therefore reduce the potential utilization of capacity. The longer term efficiency.estimates for Mozambique industry suggest that there are few major resource allocation problems within industry; however, the effects of structural adjustment on expenditure imply the need to examine more carefully the level of overall demand for industrial products vis-a-vis food, energy, transport etc.-i.e. the inter-sectoral allocation of resources, and the justification for maintaining or increasing industrial capacity levels. Thus demand projection wo-uld be a critical element in enterprise feasibility appraisals. Table 4.2: UOZAIQI -45- ENTERIS SURVEY DOMTIC RESCC COST ANALYSIS (SOT RW) CATE(3 A Comparator 2/ Short Run CIF Average aIbort aor cost ORC Maputo Capacity Tot Total Product Ratio Price Uti I at)on Cost Cost () (X) (X) (X) sic Si CaShoW Nuts 0.6 1 6 Sugar (1) 1.5 0.23/kg. 6 29 Sugar (2) 2.0 0.28/kg. 6 8a Spaghetti (1) 0.6 0.60/kg. S0 64 9 Spaghetti (2) 0.6 63 6 Crem Crackers 0.6 1.26/kg. 40 s0 9 SIC 82 Cotton Diapers 0.6 0.60 ea. 20 SIC 88 Sawn Timbr (1) 1.2 Pi 21 10 Sawn Timbr/Part. Sd. (2) 0.6 80-86 40 10 SIC 34 Cariboard from Kraft Paper 0.6 60/ton S0 76 1 Cardboard Packing Boxes 0.6 81080/ton 65 1 SIc 86 1T10x 20 SE Truck Tyro 0.5 U226/unit J8 (Actual) 67 14 Paraffin Candlox (C0 g.) 0.8 30.80/dozen 25 88 - Vinegar (OX Acetic Acid) (1.0 81.00/It 88 SIC 88 ETcFtric Transmission 81160/bs 76 74 Cable (4 specs.) 0.8 (avg) 12270/km (Prices FOB Oporto) 82960/km $3060/ks Replacement Auto Radiators 0.8 842/kg 76 81 4C Bicycle: One Sped 0.4 390/unit 75 69 5 Refrigerator (11 ft.8) 0.6 8686/unit sO 78 Nech. Water Pumps 0.8 8440/unit 60-70 70 - 3 All Ceiculatioms _sed em official eachang rate/cIr priam proevliin at the date of preparaton of pro cob data. Prealling CIF Iap to price spos which OK coleulatles w_ bed. n BS on estites from the Minitry of Agriceltur . / Note: The rensuts of thO 10-enterprise review (table 4.4) wre essumed to be more acuate than the le sintenlve 46-entrpris survey. More products are dupIlcated by both urvey , the 45-enerrise relt haw bew delee. Table 4.3: MOZASJBQE -'46 - ENTERISE S3EV1R DOIES1TC RESOURtE COST ANALYSIS (SHORT MMl) CATEGIRY B (Approuimetion*) jl Estimated Comparator V Short Run CIF Average o Dorts Labor DRC Maputo Capacity TOiW1 TotsT Product Ratio Price Utilization Cost Cost (USil) (U) (X) () SIC 32 CAM- TextiIes (1) >1.0 t0.90/yd.0 60 Cotton Textiles (2) >1.0 40 Synthetic Textiles )1.0 81.20/yd.s 83 Cotton/Spring Mattress (1.0 834/unit 80 13 12 Garments - Local Fabric )1.0 Garments - Imported Feb. (1.0 SIC 34 iWrFTlg Paper from Recycle >1.0 s0 SIC 85 37/4PVC Garden Hose 30.50/meter 78 1 2-1/2 I Polyethylene Bottle 80.40/unit 78 1 Size 88 PVC Ladies Sandals 10.90/paIr 78 1 Paint (Latex Emlsion) (1.0 82.60/1Iter 10 70 9 Paint (Industrial Enaml) (1.0 *3. 0/lIto? 10 72 7 Paint (Maritim) (1.0 88.10/lItOr 10 71 8 SIC 86 Clas ra )1.0 10.70/kg. 25 - - SIC a9 Sewing machine - cIlc. )1.0 3218/unit C0 40 - Sewing machine - m_ch. Approx. 1.0 3107/unit C0 65 - Metal Fabrication - (1.0 T0 a5 81 Aggregate Costs (Tanks, Wagons) y Is lieu of production cost emdile CIF price data to calculate RC ratio, repreeete best estlmte be wpon supplementry ouro of lforstloe (e.g. comwartl,*e Costs). y RepreenOt hpbothetical price sue that eOC ratio * 1.0 rathr thb e actual prise unleme deoted by so asterisk (O). May be usd as be_ lne refeenc for as lessig efficincy ooc aetbal CIF price i knows. move denoted by as asterisk, rupreuts an acta l reported CIF price at tie of letrliew. Table 4.4t MOZAiWiqE - ItN ENIERPRISE SURYET 1 DOWESTIC RESOMRCE COST MALVSIS (OHMIT RUN) Product Short run Comparator Averap Capacity I Share In total costo ORC CIF pric, utilization (I) Imports Labor SIC 31 Copra 0.47 8800/ton s n.a. 26 64 Margarine (avg) 0.32 81310/ton 14 59 4 Aolmal feod(avg) 0.48 25/61ton 42 39 3 Copra ol I - refin d 1.71 *825/ton 7.0 22 62 - crude 1.42 1680/ton 7 28 4 SIC 32 anvas shoes 0.37 # /pair 35.0 68 21 SIC 84 Carto- 0.93 90.60/unit 15 62 11 SIC 85 Laundry soap 0.65 OOO/ton 19 6t 4 Mediu soap 1.44 8760/ton 49 57 2 Toilet soap 1.22 S1000/ton 8 6T 2 Tires 0.80 82.29/unit 78.0 81 11 Inner tubes 1 -0.78 0.567/unit 56.0 65 5 2 2.04 80.91/unit 55.0 65 6 SIC 86 Nipj cement -Import clinker 8.20 865/ton 8.8 64 21 -local clinker 0.67 #65/ton 88.0 43 n.n Clay bricks 0.76 838.5/ton as 82 40 s0 SIC 87 FTI.-ipe 4.20 *801/ton 5 18 42 0.I. shet 0.72 $168s/ton 25 65 ? Steel pipe 8.86 $861/ton 5 10 48 St" l tube -2.20 82900/ton 7 s8 4 St" l drums 0.69 S12.6/drum 11 64 2 iBln d oe report of Arthur 0. Little for the IDA 1edustriol Eo ler,rlee RestruCtR PrOlNee. ost date deebttul. * P*rce date doebul. Source: leed s a repert bfor IDA by ArUtr S. Little, Ioo. - 54 - V. THE DEVELOPMENT OF InDUSTRIAL POLICY A. Introduction 5.01. This chapter outlines possible changes in the industrial policy framework which would be part of an overall strategy aimed at increasing output and the economic efficiency of industry in Mozambique. It is not practical to deal here in depth with the wide range of policy instruments discussed, and the purpose cof this chapter is therefore limited to considering general approaches and options rather than detailed proposals. The basis for this strategy can be interpreted as *trade efficiency' - i.e. the maximization of the gains from the use of the economy's comparative natural and resource advantages. This strategy is relevant for the planning of participation of an economy in international trade, and is implicit in the efficiency analysis of chapter 4. 5.02. It can be argued that the 'trade efficiency' approach to development policy is inconsistent with alternative development objectives, such as those focussing on domestic linkage generation, self-sufficiency, and basic needs; these latter may be termed generally as structural objectives, while the objective implicit in trade efficiency is net income maximization. A further distinct objective may be that of maximizing employment creation. However, there is in fact a close connection between the trade efficiency objective and the structural and employment objectives. The relationship between 'trade efficiency strategy' and others is as follows: 5.03. Employment generation. If a strategy is efficient in terms of maximizing the gains from trade it should lead to an increase in overall employment, both though the expansion of labor-intensive lines of production, in which low wage countries will have a comparative advantage, and through the selection of available labor-intensive technologies. An efficiency strategy will allocate foreign exchange to exporters, (although the scope for new manufactured exports is likely to be limited), and to efficient import substitute producers; both of these groups of firms, if they are economically efficient, are likely to be relatively more labor-intensive than the average for all manufacturing. Further the shift in relative prices resulting from the proposed changes in incentives should result in some shift of technology in a labor-using direction. 5.04. Maximizing domestic linkages and self-sufficiency. Domestic linkage generation could be covered initially under the trade efficiency approach by licensing policy tovards competitive imports. The aim, however, is not to maximize all possible domestic linkages, but to encourage those which appear to make economic sense. A policy of forming linkages when domestic suppliers are excessively high cost or produce poor quality can simply penalize users, and may render their products uncompetitive. However under the policy outlined above linkages that are competitive, or where there is longer-run dynamic potential, will be encouraged, both through import licensing controls and in the longer-term through special tariff prote_tion for dynamic or infant industries. _ 55 - 5.05. Producing aoods to meet basic consumer needs. It is not necessary that basic need items such as foodstuffs or essential clothing have to be produced domestically. It may make sense to concentrate resources on other goods and import certain items, even though they are necessities. 1/ The choice of what should be produced domestically cannot be determined a priori, but requires information on the comparative cost position of different goods, (such information has been calculated using DRC ratios in chapter 4). Where such essential goods have low DRCs they will appear as priorities in the trade efficiency strategy. 5.06. In general, therefore, a strategy of this type can be sufficiently flexible to meet most government objectives for industry. The proposals put forward here are therefore not part of a free-trade, solell market-oriented programme, but imply a rationalized system of protection, that involves identification of priorities for industry based on economic efficiency criteria. 5.07. This chapter considers a range of measures relating to both the short-term (up to 12 months) and the medium term (up to 5 years). This is a preliminary rather than definitive analysis, because of the complexities of particular incentive systems. Five broad areas of policy are considered: (a) Exchange Rate Policy; (b) Foreign Exchange Allocation; (c) Import Tariff Regime; (d) Controls on Domestic Prices; (e) Fiscal Incentives to Private Investment. 5.08. The policies discussed cover government interventions in the economy which may be less relevant in the lcnger-term (up to ten years) since over time changes brought about by the Economic Rehabilitation Program may reduce the need for controls over prices and foreign exchange allocation. Therefore the discussion here focusses on the next one to five years during which time interventions are likely to continue to be of importance. The aim is to identify ways of rationalizing and improving the existing system of interventions. B. Exchanae Rate Policy Objectives of Exchange Rate Reform 5.09. The official price of foreign exchange is a key macro- economic parameter, but also one which has important implications for individual 1/ There has been a special case made for food security reasons; however even here an alternative is storage and buffer stocks. - 56 - sectors like industry. As previously stated, a central part of the ERP has been the modification of the fixed exchange rate policy of the first half of the 1980's, whereby the Hetical was pegged to a basket of currencies at a rate that increasingly failed to reflect the scarcity value of additional foreign exchange. By 1986 despite being traded officially at MT40/US$ the Hetical was being bought and sold in the parallel market at a very substantial discount, so that at one point the unofficial rate is reported to have reached a ratio of 40sl relative to the official rate. 5.10 In January 1987 the fixed exchange rate policy was substituted in favor of a series of step-wise devaluations that have brought the official price of foreign exchange closer to that on the parallel market, whilst devaluing the currency substantially in real terms. Figure 5.1 sets out movements 4n the nominal official exchange rate, the real exchange rate, and domestic prices. The change in the nominal rate, adjusted for relative movements in f6reign an,d domestic prices, gives the degree to which the real exchange rate has changed over this period. This latter estimate can be only approximate given the unreliability of domestic price indices for Hozambique, but the results show a very substantial real devaluation in 1987, with a further more modest real fall in 1988 and 1989. Figure 5.1 EVOLUTION OF EXCHANGE RATE AND PRICES 1.1 | 0.4 w ~ 0.1% 7I79'/ 'I 07/_ e °,: / 0.4. ' 0.t ._ 0.1. 0.~~~~~40 4 lOM - 57 - S.11. One main objective of the exchange rate policy has been to attract foreign exchange out of the unofficial market by reducing the gap between prices in the two markets, thus increasing the amount of foreign exchange through official channels, and reduce capital flight and nonessential consumer imports. At the end of 1989 the official exchange rate was 40 to 452 of the parallel rate, and appears to be stable in that range. Figure 5.1 shows indices of nominal and effective exchange rates, and the consumer price index. Effective rates are based upon trade weights for Mozambique vie a vis principal trading partners. The nominal exchange rate, which determines imported goods prices, has increased considerably faster than the general price level, while the effective exchange rate has fallen drastically, levelling out in 1989. 5.12. A further major objective of exchange rate policy is to increase the relative price of traded goods - the goods that an economy can buy and sell on the world market - relative to those that are non-traded internationally. This price effect should stimulate production of exportables and import-substitutes with higher domestic resource content, and thus help improve the foreign trade balance. How far such production effects follow a devaluation will depend on both the size of the relative price movements - in terms of traded and non-traded prices - and the responsiveness of producers and consumers to such price changes (demand and supply elasticities). Empirical measures of the real exchange rate, such as that used in Figure 5.1, approximate the shift in traded vis-a-vis non-traded prices. 5.13. This second aspect of exchange rate policy has the most direct implication for industry in Mozambique. In most economies it is appropriate to class Manufacturing activities as internationally treded - so that most manufactures are either potentially exportable or potential substitutes for imports (subject to quality differences). Effects of currency devaluation oan industry efficiency and exports 5.14. After a devaluation direct industrial exporters, with prices fixed in foreign exchange, will find their receipts in local currency are increased, and import substitute industries will be able to charge higher domestic prices, whilst still remaining price competitive with imports, which are now more expensive with devaluation. At present in Mozambique the part of the manufacturing sector that is likely to benefit directly from devaluation is the export-oriented agro-processing branches such as cashew, sugar, cotton and tea, because they will receive a relatively large price incentive. Their local costs of production will increase due to domestic inflation, but to a lesser extent than border prices in devalued meticals. The effects on other export manufacturers are difficult to assess because they have exported little in the past apart from small amounts of cement, tires, batteries and various metal products. In 1987, for example, out of total national exports of US$ 98 million agro industries - cashew, cotton fibre, sugar and cashew oil totalled US$43 million. There were some miscellanec." manufactures in the general category of 'other exports' but - 58 - in total this category was less than US$8 million. 21 5.15. This is not to suggest that no new manufactured exports will be generated by the major changes in the exchange rate. The DRC estimates from the survey of enterprises indicate that a broad range of goods are competitive internationally; however, the long run position is difficult to determine in the current fluid situation. Given an improvement in local infrastructure and the development of appropriate marketing links some new exports by such activities should be feasible. However an important general point is that where a strong industrial base with the necessary support services for exporters is lacking, a policy of changing the exchange rate is, per se, unlikely to stimulate a significant export-led expansion of manufactures. Experience elsewhere in Africa bears this out, since despite devaluation in many countries growth of manufactured exports from the continent has been generally disappointing. 3/ For much of the manufacturing sector in Mozambique, at least in the medium-term, exports may remain a residual relatively small outlet, with the majority of manufacturers continuing to supply the domestic market with various types of import-substitute production. Equilibrium Exchange Rate 5.16. The future policy on the exchange rate is outside the scope of this study. However the divergence that still exists between the official and the parallel rate means that the official rate is not equal to the equilibrium rate, even with the current level of inward capital transfers. 4/ The parallel rate is also unlikely to accurately reflect the equilibrium rate because it is likely to be inflated by risk premia, capital flight, and inflated profits attainable on resale of luxury goods. Under the Government's Policy Framework Paper, it is expected that the official rate will move towards a level that is closer to an equilibrium exchange rate, and will then be held constant in real terms by adjusting the exchange rate periodically to allow for movements in domestic relative to world prices. 2/ Source; Table 8.4 Informacao Estatistica National Planning Commission. 1987. 3/ Annual real growth of manufactured exports from Sub-Saharan Africa in the 1980's has been as follows: 1980-84 2.92, 1985 12.5Z, 1986 2.32, and 1987 -0.3?; (Sourcet World Bank World Development Report 1988.) 4/ The equilibrium rate is theoretically difficult to define because of the presence of inward capital transfers. It is possible to distinguish between a short-run equilibrium rate, based on temporary transfers, and a long-run rate, where transfers are at a stable and desired level over a full economic cycle. Since it is currently difficult to estimate what the longer-run average level of transfers to Mozambique will be, it is difficult to operationalize the concept of a long-run and sustainable equilibrium exchange rate. Any estimate of a short-run equilibrium rate will be sensitive to changes in capital transfers. - 59 - 5.17. Although there is considerably uncertainty regarding the equilibrium level of the exchange rate, for the present capital transfers received by Mozambique, it is likely to be close to midway between the official and the parallel rates, or in the range 1.3 to 1.7 times the official rate as of early 1989. In later sections of this chapter. concerned with tariffs, licencing and price setting, an approximate value for the equilibrium rate is required. For illustrative purposes a hypothetical rate of 302 over official rate is assumed. However, this is not intended as a proposal,, and in any event the relevant rate at any given time would vary with changes in exchange rate policy. 5/ C. Foreign Exchange Allocation Channels of Foreign Exchange Provision 5.18. For many manufacturers the availability of foreign exchange may be of even greater concern that its price, since cost increases of imported inputs can be passed on through the pricing system. At present there are five main official channels for manufacturers to obtain foreign exchange, apart from the highly restricted availability through the regular export and service earnings. (a) The Export Retention Scheme - through which exporters have the right of access to P proportion of the value of their export sales. (b) Allocations under donor import support programmes coordinated by the UCPI, including the World Bank rehabilitation credits. (c) Special 'funds' for designated enterprises, such as the *market fund' for registered traders and the 'SME fund' for producers. (d) Specific donor projects for individual enterprises and programmes. (e) The System for Non-administered Allocation of Foreign Exchange. 5.19. Foreign exchange allocation is being gradually opened up to market forces through the system for non-administrative allocation, partly funded under the World Bank Third Rehabilitation Credit. The system is initially confined to spare parts for transportation, and inputs for the garment and shoe industry and in its first year is expected to cover approximately 3Z of total imports (about US$30 million), with allocation being based on price (including possible tariffs) and not on quotas. Initially an allocation ceiling per applicant would be in force until the foreign exchange demand position can be accurately assessed. The main sources of foreign exchange are 5/ In practice, an estimate of appropriate value of the exchange rate is essential if tariffs are to have a rational basis in terms of domestic and %vorld prices. - 60 - currently external grants and soft loans (emergency and nonemergency), from donors, which account for about 87Z, and commodity exports about 102, (including the export retention scheme which accounts for about 7Z and the SME and market funds for 32 in 1989). Thus categories (b) and (d) comprise the predominant source of foreign resources for industry. 5.20. Prior to the eventual opening up of the market it is desirable that the allocation system should be made both as responsive to the needs of enterprises as possible, whilst using simple, but appropria.e, economic criteria as the basis for allocation between activities. The discussion in this section focusses on the method of allocation for the first three sources of foreign exchange identified above. The fourth - donor support to individual projects - it is assumed would be subject to assessment by the donor agencies concerned. The Export Retention Scheme 5.21. This scheme gives exporters the right to access to a proportion of the foreign exchange value of their exports. The exact proportion involved is negotiated case-by-case: initially retention ites were from 30Z to lOOX, varying with the foreign exchange content of peoduction and the perceived economic importance of the goods concerned. Contracts between the government and exporters under this scheme are negotiated annually, and all significant exporters are now covered. At present, the range of retention rates offered to different exporters has been reduced with an average of 402 to 50X. 5.22. The scheme was introduced initially as a means of ensuring a flow of foreign exchange to exporters to allow continued production, and some rehabilitation; prior to the major donor support programmes it was the main source of foreign exchange for many firms. From a short run perspective the Export tetention scheme serves an important mobilizing function. The gains from the scheme are clear provided that: (a) enterprises use retention rights to directly generate additional export revenue, and (b) the system does not encourage abuses such as over-invoicing of imports. In the longer run however it creates certain economic distortions. As foreign exchange becomes more freely available the scheme would in any case become redundant. 5.23. Problems associated with the scheme from the point of view of economic efficiency are as follows: (a) As long as the official exchange rate continues to undervalue foreign exchange, guaranteed access through the retention scheme is a form of subsidy to exporters which is not available to import-substitute producers who save but do not directly generate foreign exchange, even if they are economically efficient, or could be potential exporters; and (b) Within the export sector it is the producers with higher import component of costs who receive a higher retention rate, and thus a higher rate of subsidy. However ;uch exporters will have a weaker net foreign exchange effect than those with a higher local content in their production. The scheme could thus encourage more import-intensive export production. - 61 - 5.24. Despite the longer term issues, in the short to medium term there is a case for retaining the existing scheme on two grounds. First, it gives security to the existing export sector. With the expansion of exports a revolving fund of foreign exchange can be created, with the allocations under the Retention scheme leading to additional exports, which in turn allow both more imports for existing exporters, as well as a fund of foreign exchange for the economy as a whole. Second, provided that there are no subsidies to exports, by definition all exporters will be internationally competitive and will therefore be economically efficient. They would thus probably be among the priority category of enterprises. 5.25. For the above reasons continuation of the scheme is justified on economic grounds, although certain modifications might be worth considering, relating to (a) the question of whether the retention accounts of exporters be denominated in local or foreign currency, and (b) the percentage of export earnings that can be retained in the Retention account. (These issues are addressed further in chapter 6. Donor Import Allocation Procedure 5.26. Foreign exchange allocations from donor funds are critical to many enterprises and form the bulk of the foreign exchange available to nonexporting enterprises. Allocations are made several times a year under each donor program, and are coordinated by the UCPI 6/ in discussions with Hinistries. After an overall sectoral division has been made, Ministries identify their priorities and an attempt is made to match needs with the foreign exchange available, much of which is tied to specific country sources of supply. The criteria to be used in these allocations have been the subject of some discussion. At present qualitative criteria are applied, although the need to develop a simple but workable quantitative measure is recognized. A version of the DRC ratio efficiency measure of chapter 4 has been estimated for the main productive activities of the economy. However the results of these calculations have not been used, and looser criteria still apply. 5.27. Foreign exchange allocations by Ministries are assessed on the basis of priority needs. MIE has two priority categories: (a) around 20 agricultural and industrial intermediate inputs which are in their first priority category, and, (b) in the light industry category basic foodstuffs and textiles. However, the priorities are not applied rigidly to the exclusion of other activities, and there is an attempt to match general needs with foreign exchange availability, even where enterprises are not in the main priority group. 5.28. UCPI allocation criteria include (a) the 'relative importance" of enterprises, (b) the degree of urgency" for foreign exchange (c) relative contribution to exports, (d) relative contribution to offering essential g,ods to the population, and (e) degree of complementarily in the production process of essential commoditiest. Criteria c), d) and e) favor domestic 6/ Coordination Unit for Import Programs. - 62 - resource based and export production, and are more specific than a) and b). However, there is no assessment of the relative economic efficiency of various activities, which would ensure that the limited foreign exchange available can be distributed according to income maximization criteria. 5.29. Table 5.2 shows the allocation of foreign exchange by manufacturing branch as of December 1989 for MIE enterprises. All but nine enterprises received an allocation. The table shows that those subsectors receiving the largest foreign exchange allocation in relation to production were footwear, cooking oil, rubber products, construction machinery, chemicals, metal products and heavy engineering. On the other hand the lowest allocations in relation to production were for food, drink and tobacco. This pattern of allocation does not necessarily reflect the most efficient use of foreign exchange resources. Within the metal products category, for example, a metal pipes plant received an allocation of US$6.8 million. Yet, according to DRC ratio information, this industry is not economically efficient in the short run. Table S.1 MOWZMBIQUE - FOREIGN EXCHANGE ALLOCATION AND PRODUCTION MIS ENTERPRISES 1989 Share in allocation Share in Production (as of December 1989) (constant 1985 prices) US$million X 1989? Food, tobacco 12.2 10.9 16.2 Drink 7.5 6.7 21.3 Footwear 6.0 5.4 1.9 Light engineering 5.5 4.9 8.7 Furniture 0.3 0.3 0.6 Textiles, clothing 23.0 20.4 23.4 Constr. machinery 6.3a/ 5.6 4.2 Chemicals 9.8 8.7 5.5 Electrical goods 0.8 0.7 0.2 Metal products 10.7 9.5 1.9 Refrigeration eqpt. 4.2 3.8 3.1 Heavy engineering 2.9 2.6 1.2 Paper products 0.0 0.0 0.3 Rubber products 4.9 4.4 2.5 Glass 0.3 0.2 0.5 Cooking oils 12.1 10.8 3.9 Plastic products 5.7 5.1 4.6 112.4 bi 100.0 100.0 al Includes US$4 million of Swedish aid not allocated directly to MIS. _/ Excludes electricity, petroleum, and salt Industries. Sources MIE - 63 - 5.30. The use of an efficiency measure such as the DRC ratio set out in chapter 4 might be expected to channel the allocation of scarce foreign exchange to the those activities which will make the best use of it, on the basis of the relative efficiency of different branches of manufacturing. Several sector Ministries have experimented with this approach, but it has not been systematically introduced. Since any general efficiency measure is an approximation and an aggregated measure for a whole industry or branch may not be accurate for an individual firms. the exact allocation needs further information on the cost position of individual enterprises. A three-stage allocation procedure may be used. (a) Initial branch allocations within manufacturing would be based on short-run DRC ratios. (b) When an allocation has been agreed for a branch, all enterprises within the branch should be given the opportunity to request to take up this allocation; and (c) If any firms in the branch do not do so within a specified period - say three months - this can be taken as evidence of their high cost position. If the total requests come to less than their branch allocation the unused funds can be reallocated to other branches. 5.31. Actual allocations will rarely be as simple as this since the full amount of foreign exchange available for a year may not be known at the beginning of the planning period, and there is the further complication of tied donor aid. However the general procedure of allocating centrally on the basis of efficiency criteria, and then adjusting these central allocations in the light of demand at the enterprises level, is worth attempting to implement systematically. It should be noted however that this modified approach to allocation still requires administrative action and may be inflexible (e.g. with. regard to new startup enterprises). It is an interim measure to introduce more efficient resource allocation, complementing continued reform of prices and exchange rates. 7/ Guidelines for Donor Import Allocation System 5.32. Using the DRC measure, table 5.3 sets out a method for foreign exchange allocation via priority groupings of enterprises. This is illustrative and in practice a less mechanical approach would be justified. 7/ The move to a nonadministered system is under way pari passu with movements towards a long run exchange rate equilibrium when it would be possible to remove quantitative controls on foreign exchange allocation. In the case of enterprise-specific donor import support however, since the foreign exchange allocation is effectively administered (and probably tied) at source and is not part of a 'pool" of free foreign exchange, a central allocation system of some type would probably always be needed. _ 64 - Group 1 contains activities that on average are clearly efficient (DRC less than 1.0). This is the first priority. Group 2 may cover branches with DRC ratios of, for example, 1.0 - 1.5. The efficiency of these branches depends on the economic valuation of foreign exchange, and they would be efficient provided that the realistic exchange rate is not more taan 50? above the official rate. However, at the illustrative benchmark of 302 over current exchAnge rate those with ratios of 1.3 _ 1.5 would be marginal. 8/ This is the second priority. Group 3 branches will be unambiguously inefficient however, with DRC ratio greater than 1.5, since the economic value of foreign exchange is assumed to be less than 502 above the current rate. Finally group 4 branches would be those which have a negative net foreign exchange effect, since at world prices their output is less valuable than their traded inputs. Such cases will be rare, but this category is the extreme example of economic inefficiency and would justify closure since its continued operation would reduce the productive resources of the economy. 9/ 5.33. Under this system Ministries and enterprises could provide approximate estimates of their foreign exchange requirements at the beginning of the planning period. When these estimates are known the full requirements of group 1 branches would be met, followed by group 2. Where enterprises within a particular branch are unable to take up all of the branch allocation within a specified tize funds should be reallocated. Branches in group 4 would not receive any allocation, but group 3 enterprises may be considered as possible, judged on a case-by-case consideration of the efficiency and longer run cost position of the enterprise. A world price data bank could be maintained by the UCPI or other agency in order to facilitate the system. This would contain updated information on import and export prices for a range of important products. 8/ The illustrative benchmark is referred to in 5.17. The significance of the exchange rate is explained in note 9 of chapter 4. 9/ This applies to the short-run situation in the case of inner tube production at a sales price of $0.78 per unit. At this price the plant should be closed. However, the price assumption has to be carefully checked. (See table 4.4.) _ 65 - Table 5.2: NOWINBIQUE - ILLUSTRATIVE GROUPING OF BRANCHES BY DRC RESULTS a/ Comments Group 1 DRC <1.0 Economically efficient. Group 2 DRC cl.5 Efficiency depends on valuation of foreign exchange. al Group 3 DRC >1.5 Likely to be economically inefficient. b/ Group 4 DRC negative Unambiguously inefficient- net foreign exchange loser. C/ DRCs calculated valuing foreign exchange at official exchange rate. b/ It is assumed that the appropriate (real) exchange rate is less than 502 above the official exchange rate. The Small-Medium Enterprise and Market Fund 5.34. These funds disbursed approximately US$12.5 million in 1988 and US$25 million in 1989. In 1988 180 small and medium enterprises were identified as eligible to participate in the fund, as a result of consultations between the Ministry of Trade and Provincial Directors of industry, of which 81 received allocations. Foreign exchange is allocated on the basis of need, and the government will define the basic range of goods that can be imported under the fund, without specifying which products firms should import. Enterprises in 1988 applied for allocations of between US$3,000 and US$50,000 up to 4 times a year. The selection of the 80 enterprises is administrative. However since the size of allocations averages less than $100,000 per annum per enterprise, the use of efficiency criteria is less important. It seems adequate at the this stage to monitor the general financial and technical performance of enterprises participating in the scheme utilizing contacts at the provincial level. Import Licenses 5.35. After an enterprise has been allocated foreign exchange under these various mechanisms an import licence is required from the Ministry of Trade to allow goods to be imported. The process of obtaining licenses has been accelerated, although enterprises still complain about delays (largely due to foreign exchange scarcity rather than administrative problems). 10/. On efficiency grounds there is a case for substituting a system of Import protection based on licensing controls for one that regulates the demand for imports through tariffs, or exchange rate policy. The move towards use of prices and tariffs as the sole protective mechanism in Mozambique is under way, bit will take some time, requiring particularly a further normalization 10/ This is discussed in Report on a Feasibility Study for Export Processing Zones in Mozambique, op cit, 1989. - 66 - of the foreign exchange position. In the interim certain measures may be taken to assist in the transition to price allocation. 5.36. Import licenses are not normally granted for goods which compete directly with domestic products (although this reportedly did occur in 1988 in the case of garments and some metal products, and there have also been emergency program imports such as blankets which have competed, as discussed in chapter 3). Current policy in principle is that import licenses will not normally be issued when domestic alternatives are available at the time of the import request. Examples are licenses for steel, paint and asbestos, that were turned down in 1988 on the grounds of the availability of domestic supplies. Before an import licence can be issued a letter from the domestic supplier is required stating that it is not in a position to meet the order. 5.37. This form of protection through licensing could be justified as a means of fostering domestic inter-industry linkages and protecting new producers at an early stage of their development. However experience with this policy in other countries has shown that it can have significant short-term costs, that in some cases may outweigh any longer-run benefits. The important general point is that just because a good can be produced domestically does not mean that trade in competing goods should not be allowed. The case for this form of protection is strongest where the protected domestic industry, although currently uncompetitive, has the potential for cost reductions and technical progress; in other words it can be classed as a 'dynamic" or 'infant' industry. However where such potential does not exist this form of protection will impose a short-run cost on users and consumers, without any prospect of longer-term economic benefits. The normal justification is in terms of the protection of jobs and this social justification must be balanced against the economic costs which can be particularly significant if users of the protected good are competing in the international market as exporters. 5.38. In order to address this issue . the policy of non-issuance of licenses to imports that are competitive with domestic goods could be modified to include: (a) the incorporation of price and quality comparisons between domestic and foreign goods; and (b) an assessment of the potential of the domestic import-competing producers for longer-run cost reductions and improved operating efficiency. Revised Guidelines for Import Licenses 5.39. Under this revised scheme imports would be allowed if the local substitute product is of a sufficiently high price or low technical standard in relation to imports to impose serious costs on users. However where a dynamic potential can be demonstrated, protection could be continued despite a current cost or quality disadvantage, on the grounds that the local producer needs to get established in the home market before it can achieve its full potential. - 67 _ 5.40. Some guidelines are required to make these general recommendations operational. It is not appropriate, until prices (with or without import tariffs). are fully adjusted, to expect that local goods be price competitive with imports. It is suggested that therefore that a Omargin of domestic price preference' is established, allowing domestic er-factory prices to exceed import prices at the official exchange rate by up to the margin, before imports are allowed. The margin should be a function of the difference between the parallel and official exchange rate. The illustrative example of section 5.17 used a 302 margin; however whatever margin is employed, it would be reduced pari passu with developments in exchange rate policy, and further real devaluations. In this sense it would have the same effect as a tariff; it would therefore be relatively easy to move from this interim position to a tariff based system at short notice. 5.41. In the illustrative case here therefore imports could be allowed if domestic prices exceed import prices at the official exchange rate by more than 30Z. However quality considerations should also be taken into account, since if a local good does not meet required minimum technical specifications it will not be an adequate substitute for imports. 11/ This justification for allowing imports in the face of domestic competition would have to be examined on a case-by-case basis drawing on relevant technical evidence. Under exceptional circumstances imports may be restricted even where domestic prices are high and quality of lower standard than imports, in order to protect a local producer with potential for longer term efficiency, or alternatively to provide interim protection for jobs. Sumary of Recommendation on Foreign Exchange Allocation 5.42. A summary of recommendations on foreign exchange allocation to industry is as follows. (1) The retention scheme is justified in economic terms as a medium term instrument. However it may be desirable to clarify certain aspects of the scheme, possibly denominating accounts under the scheme in foreign currency rather than in Meticals. (See further chapter 6). (2) An efficiency measure (e.g. DRC ratio) should be used as a guide in allocating foreign exchange under general import support programmes, as an interim measure complementing continued price and exchange rate reform. Two categories of producer could be established within manufacturing - with category 1 having priority in allocation, and category 2 having the bulk of the remaining foreign exchange. Individual firms not within these categories could still receive foreign exchange, but a special case for their allocation would have to be made by their 11/ The Ministry of Trade has used this criterion. for example in relation to imports of new types of cans which Metal Box cannot supply, and cartons which CARMOC cannot provide. _ 68 - sponsoring Ministries. DRC estimates would hbve to be updated and a world price data bank set up; and (3) The policy on the non-issue of import licenses when domestic substitutes are available should be modified. Imports should be allowed if the price of the competing domestic good is above a certain "margin of domestic preference". This would be a function of the margin between parallel and the official exchange rate. A margin of 30Z is used here for illustrative purposes. Alternatively imports may be allowed where the quality of domestic production is so low that it is not an adequate substitute for imports. Where a local good is at high costiquality but is judged to have a potential for significant cost reduction, a special case could be made to restrict imports, in order to encourage domestic production. D. Trade Tariff Reform The Current Reform Program 5.43. Trade tariffs are an important instrument of industrial policy. However, up to now the Government has not used tariffs as a tool of economic policy, but more as a limited revenue raising instrument. Imports are limited predominantly through quantitative restrictions in the form of import licenses or foreign exchange allocation limits, with tariffs playing a minor role. 5.44. At present the government is implementing a two-stage reform of the tariff structure with a view to adopting standard international (Brussels) nomenclature for tariff headings, converting all specific rates to an ad valorem basis, and eliminating a range of minor trade taxes. At a second stage the rate structure would be simplified by reducing the number of product categories subject to separate rates, and by reducing the range of rates. In addition at this second stage questions of the wider role of the tariff in generating revenue and protecting local producers will be considered. The revised tariff structure consists of several bands of rates, 1- 12? in the case of basic foodstuffs, 50 -632 in the case of luxuries, and 15 - 402 for others, and in some cases two rates are in force for the same product, depending on whether it is purchased for production (lower rate), or distribution (higher rate). Exporters are exempt from duties if they produce with over 35? local value added. Producers for the domestic market incur a reduced rate if they produce with over 452 local value added. There is no formal duty drawback system. There is also an export tax, currently of 52 applicable to all products. 121 121 The rationale for this is to maximize foreign exchange revenue to the country, which presupposes that the products in question face inelastic export demand. - 69 - 5.45. Since the first stage of the reform is currently underway, as set out within the Government's Polic7 Framework Paper, and since it has also been considered in a recent IMP study on taxation in Mozambique, the discussion here is limited to how, in the medium-term and beyond, tariffs can in principle be used as protective measures to encourage relatively efficient industrial development in Mozambique. 13/ 5.46. At present, import licensing is the main form of protection, and there is a case for moving towards a system of protection in which tariffs play the main role, along with continuing price and exchange rate movements. Under normal trading circumstances tariffs have two main advantages over quantitative controls. (a) They bring in revenue to governments, whilst normally under an import control system scarcity margins on imports are captured by traders or direct importers. (b) They provide a protective margin whilst not insulating domestic producers entirely from price movements on the world market. The reforms discussed here will become relevant once import licensing ceases to be the main form of protection and the demand for imports can be by controlled through the price system, an objective which is being supported by the IMF and World Bank. 141 5.47. This discussion of future tariff reform must be somewhat speculative, since the full implications of the suggestions made here would have to be worked out in detail at a later stage. The proposal is for a tariff framework based primarily on the need for a rational system of protection of the industrial sector. However turiffs, like other forms of taxation always contribute to more than one policy objective. It is not possible to consider a future tariff structure from the perspective of protection alone; revenue and distributional equity objectives will also be involved, and an attempt is i,-de to show how these can be fitted into the system. Tax Revenue Considerations 5.48. There is a strong case for raising additional revenue from taxes on trade. In Mozambique at present trade taxes actually paid are low, partly because of a widespread system of exemptions; the IMF estimates that prior to the first stage of the tariff reform 602 to 702 of imports paid no import duties, although the percentage of imports exempt from customs fees (EGA) was 13/ The reform of trade tariffs alone is not sufficient to establish an efficient structure of taxation, but has to be linked to the domestic tax and incentive system; otherwise attempts to rationalize tariffs may be nullified. A specific problem of this type may for example exist with the "imposto de circulacao' (turnover tax) which creates distortions because it accumulates at each stage of production. (See para 5.100). The incentive effect of domestic taxation is an important area for further study. 14/ For example under the terms of the Third Rehabilitation Credit. - 70 - lower. 15/ In 1986 the average effective tariff rate (total duties, fees and taxes collected on trade to total imports) was around 7.SZ, which .s low by the standards of most developing countries. The consequence has been that trade taxes are providing a much lower share of total tax revenue than in other countries. This is illustrated in table 5.3 where the position in 1988 is compared with that for other African economies. Table 5.3t MOZAMBIQUE - TRADE TAXES IN TOTAL TAX REVENUEs MOZAWBIQUE AND OTER AFRICAN SCONOMIES 1980s Average for other Mozambique bl African Countries (early 1980's) (1988) Tax Revenue (2 of GDP) 16.1 16.9 Tax Structure (2 of Tax Revenue) Taxes on income and profit 29.5 29.0 Taxes on domestic commodities 27.9 53.3 Taxes on international trade 42.6 17.7 Sources a/ Anderson (1987) table 2.1 p.5. k_ IHF (1988) table 2 p.61. Designing the Tariff Structure 5.49. The opportunity exists to avoid making mistakes that have occurred in other countries as a result of the construction of complicated and often self defeating systems of protection (e.g. with offsetting tariffs and subsidies). This is because tariffs have not up to now played an important role and so there are no vested interests in industry, commerce etc, in maintaining specific tariffs or subsidies in order to protect particular industries. Starting from a relatively 'clean sheet" it is possible to set up a reasonably optimal system which has the advantages of a) consistency, and b) transparency. 5.50. In devising a tariff structure it is simplest administratively and efficient in terms of resource allocation to set an initial standard uniform rate of customs duty which will be imposed on all dutiable goods - unless these are granted specific exemptions. This uniform rate would be set according to a) exchange rate targets, and b) revenue targets, allowing for likely levels of exemption. once this rate has been agreed various deviations from it could be allowed to meet other objectives e.g a) special 15/ Memorandum on the Tax system of Mozambique IMF, March 1988. - 71 - deviations from it could be allowed to meet other objectives e.g a) special protection and b) distributional equity. An DMF proposal is that the custom fees (EGA) currently imposed at a 10 flat rate, should be merged with customs duties. If this proposal is adopted this standard uniform rate for revenue purposes could incorporate the EGA. 16/ 5.51. In practice there would always be some tariff exemption for goods that fall into special categories, but from a revenue, resource allocation and incentive point of view these should be minimized. Donor support programmes may insist that particular goods are tax exempt, particularly for items imported as emergency relief, or as gifts. However, exemptions under the investment incentive legislation are under the government's control and can be reduced. This is argued below, where it is suggested that tariff exemption on capital goods imported for new projects under the incentive legislation should be reduced and ultimately ended. 5.52. The basic uniform rate of customs duty should be related to the margin of preference (See 5.40), which for illustration may be assumed at 302, in line with the parameter suggested in section 5.17, as well as to revenue requirements taking account the levels of exemption that remain. Apart from administrative simplicity a single uniform rate of tariff has the advantage that its effect on the incentive to produce is neutral as between activities. In this context the concept of the *effective protection rate' (EPR) is useful. 17/ With a single uniform tariff, and no exemptions, all producers will nave the same EPR which will equal the single tariff rate. Thus there will be no special incentives from the protective system for particular producers. 5.53. A major objective of a tariff reform should be to ensure that the range of EPRs generated by the new tariff structure is in line with 161 However, if it is decided to retain the EGA as a separate charge, the minimum rate would be shown as a combination of two separate charges, one for EGA and the other for the standard customs duty. Further, exemptions for raw material and intermediate good imports should be limited to exporters alone. Export costs are raised by any taxes on their inputs and import tax exemptions for exporters are a recognized incentive that does not contravene GATT1. 17/ The EPR measures the extent to which value-added in an activity is increased by protection in comparison with what it would be with a free trade or zero tariff policy. The EPR can be defined as: (value added with protection - value added without protection) x 10OO value-added without protection Strictly it is the average effective protection rate, and not the nominal uniform rate, that should be related to the 302 "margin of domestic preference". However, it may be assumed that Lhe EPR for a range of industries will vary around the average nominal rate. - 72 - government objectives in terms of encouraging priority activities, particularly dynamic or infant industries. Therefore, it is necessary to try to construct a tariff system which will generate broadly the desired range of EPRs. This can be done by adding additional layers of tariff rates in line with the objectives of industrial promotion. In addition there will be ;distributional equity considerations which will influence tariff rates. These are discussed before returning to the issue of determining special protective rates. Income Distribution and Tariff Policy; Luxury Goods 5.54. On distributional equity grounds there will be some categories of imports which the,government is likely to wish to tax significantly more highly than the average, because they are nonessentials, or luxuries. Further where their demand is price inelastic they may also be significant sources of tax revenue. 18/ Supplementary taxation of luxuries is commonly carried out through either a) the trade tariff system. or b) the excise tax system (imposto de consumo'). Using tariffs, for this type of import there could be a relatively high additional rate of import tariff, for example 302, which would be on top of the basic flat rate of 302. If these rates were adopted such goods would therefore have a total import tariff of 60!, composed of the basic rate plus a premium rate for luxuries. There is not a strong case for differentiating between different types of luxury imports - so that a single rate of tariff ;an be applied for all goods that are classed in this category. If the Oimposto de consumo- is used then the excise tax rate would be 30! of the CIF import value of the luxury product in question. 5.55. Using the tariff approach, a high tariff on "luxury* consumer imports could provide an incentive to produce these locally, because the tariff will create a high EPR for local production. (I.e a local producer would be able to sell at a price well above the import parity price (cif price at official exchange rate) and still undercut imports because of this additional tariff protection). To avoid creating incentives for the local production of luxury goods it is necessary to offset the effect of their additional protection. This can be done by imposing a domestic excise tax on any domestic production of these goods. This tax will be collected from domestic producers and should be set at a rate that equals the additional tariff on luxury imports. Using the tariff rates mentioned earlier, if the additional tariff rate on luxury imports is 30! then the offsetting excise tax should be set so that it generates revenue equal to 302 of the cif price. 5.56. For example, in the case of televisions, if the cif price is US$300, and extra tariff protection is 30! (in addition to the basic 30Z), then there should be an offsetting ad valorem domestic excise tax on local 18/ A luxury good may be defined as a good for which consumption expenditure increases more than proportionately with an increase in income. In the Mozambican context likely candidates are higher priced clothing, domestic electrical appliances, spirituous liquor, and passenger vehicles. _ 73 - television production set to generate a revenue equivalent to US$90. Therefore if there is any existing local production of televisions the percentage rate of excise duty will be calculated as US$90 at the official exchange rate divided by the local ex-factory selling price of televisions. Periodically this rate of excise duty would need to be reviewed in the light of changes in tariff pclicy and in the relation between the domestic and import price of televisions. However where local production has not yet started, it would be necessary to make clear that in the future any local television production wc'uld be subject to an excise tax calculated in the manner outlined. In this way the additional import tariff of 30? of USS30 will not protect local television producers. 5.57. If the -impost6 de consumo excise tax method is used in order to deter consumption of luxury goods, there would not be any need for the complication of a special domestic tax offsetting an import tariff because the excise tax would be levied automatically on both imports and domestic products. In terms of the example here, goods designated luxuries would carry an imposto de consumo of 302 whether they are produced domestically or abroad. In terms of simplifying the tariff system this would be a good alternative approach. It would also mean that the objective of trade protection is treated separately from the objective of distributional equity, which may be regarded as more rational in policy terms. However, an important consideration is that of efficient tax collection. Since the efficiency of tax collecting at the border is usually greater than it is once goods get into the marketplace, there may be good grounds to favor the tariff approach, both in terms of revenue collection and effective policy. 19/ Tariffs and Long-Run Protection 5.58. The type of protective tariff recommended here differs from that employed in a number of countries where the rate of tariff protection is determined by the end-use of a good; (e.g. final consumer goods are commonly given significantly higher rates of tariff protection than intermediate and. capital goods). The end-use approach is reflected in the recent tariff structure announced by the Government, in which imports for consumptirn receive higher rates than imports of the same goods for production. The problem with this type of protective system is that it can generate very high EPR's for consumer goods irdustries which may be neither economically viable nor produce the type of commodities on which the government places the most priority. The EPR is high because a consumer goods enterprise would have lower rates of tariff on its inputs of intermediates than on its outputs of final goods. 19/ Because in terms of an effective price incentive it would, be better to have a second best system that works properly than a first best system that does not. _ 74 - 5.59. It is justifiable that selected activities should receive special encouragement from the tariff system. These would fall into two broad categoriess (a) dynamic or infant industries with a potential for long-run efficiency provided that new investment or rehabilitation is undertaken; (b) inefficient industries with little dynamic potential but which the government wishes to support because of their employment effects. In Mozambique currently' rehabilitation is likely to be a more relevant concern than infant industries; temporary protection may be justified for enterprises considered to have significant cost reduction/productivity improvement potential. This may be restricted to enterprise which are implementing rehabilitation programs or have submitted them for financing. The size of the second group will depend upon the degree of resource immobility internally - i.e. alternative opportunities for employment. Any industries requiring special tariff protection must make a case to the government; any protection granted should be reviewed periodically to assess the progress of change and restructuring in the protected activities. 5.60. It is recommended that a permanent Tariff Commission be established within the Government to deal with requests for this type of special protection. This commission could request information from enterprises and their sponsoring Ministries for example relating to: - costs of production - size of expected market - productivity performance over time and gains from rehabilitation - technical capability - prices of competing imports - employment effects 5.61. Estimates of the economic viability of producers would be needed, supplemented by plans for further cost saving and productivity gains. For the second category of producers, to be protected for employment reasons, there would also need to be a plan setting out proposals for phasing out of unprofitable/inefficient production, with alternative proposals including possible product diversification. 5.62. Where enterprises are judged to qualify for special protection under either heading they could be granted an additional protective tariff on competing imports. Since the objective is to grant a. special incentive to their domestic production there should be no offsetting domestic excise tax on production. These Omade-to-measure" tariffs would just allow domestic prices sufficient to earn a non-monopolistic rate of profit. however, in practice, calculation of such specific rates can be complex. A simpler approach is to use a small number of additional tariff rates for protective purposes, which will be in addition to the standard uniform rate. For example, with the standard rate of 302, one could add three possible protective rates of 102, 20? and 30Z. Enterprises judged to qualify for _ 75 - special protection would be given one of these three protective rates varying with their competitive position in relation to imports. Thtt weaker producers would qualify for the higher protective rate. 5.63. If producers receive these additional rates nominal protection on their output would in this illustration be 302 plus the special additional rate so that their EPR should be raised significantly above the standard nominal rate of 30Z. However it should be noted that once this differential set of tariff rates is introduced it becomes more difficult to predict EPR's. Some producers who receive special protection may also use protected inputs that a.so receive special protection, and this will partly or wholly offset the gain from the protective tariff on outputs. To avoid a general problem of offsetting input ta-iffs special protection should only be granted to a small number of producers. 5.64. A weakness of this type of protective strategy is that by sheltering producers from import competition in the short-run, it may give them insufficient incentive to lower costs over time. To ensure that the system is adaptive a phased timetable of reductions in special rates of protection could be considered. This might operate by granting the maximum special rate, e.g. 302, for the-first three to five years, which would then change to the second rate, e.g. 202, for a further three years - falling to the third rate of 102 at the end of this period. Over time if approximate international competitiveness is achieved the normal basic rate of tariff would become applicable. Export Promotion and Tariff Policy 5.65. A tariff system should be designed so that it does not ienalize the export sector. Exporters sell at world prices, so that unless some form of export subs.(dy =cheme is. in operation an export enterprise is not protected. Howevex, if imported inputs used by exporters are taxed then the exporters may face negative effective protection, since whilst their output is sold at world prices their inputs will be increased above world levels by the rate of tariff. (In fact under the latest tariff rules exports are entitled to duty rebates subject to local value added requirements). 5.66. The objective of tariff policy in relation to exports should be to try to avoid such negative protection. In the case of imported inputs used in export production this can be partly ensured by a temporary import regime using a duty-drawback scheme, in which exporters are refunded the tariffs on the imported inputs. However, negative. EPR can also arise if exporters purchase inputs from domestic suppliers, when these suppliers are protected by the tariff system and therefore can set their own prices at above world levels. 5.67. The desirability of granting exporters access to inputs at world prices is widely accepted as an export promotion strategy. 20/ Under the type of protective system outlined here, purchases by exporters of protected 20/ Tax incentives for exporters of this type are also acceptable under GATT rules. - 76 - domestic inputs can be made equivalent to purchases at world prices by granting a credit to exporters equal to the difference between the domestic and world prick value of the inputs involved. The incentive could be granted through a tax credit. 211 5.68. There are various ways such a credit could be paid; a) deductible from company A's profits tax liability, b) separate cash payment for company A if it is incurring (short twirm) losses; c) applied to the cost of Company B's product. The credit could be awarded at the beginning of each tax year, anr; lculated in relation to purchases in the previous year. It would be up to exporting enterprises themselves to claim their right to such a subsidy, with the merits of the case determined by the Tariff Commission. This is one case where there is no easy alternative to using a subsidy to offset a tax. However, the source of finance for such a credit to exporters would have to be carefully examined, and in allocative efficiency terms the preference would remain one of longer term exchange rate reform rather than complicated fiscal arrangements. Summary of Tariff Refom Recommendations 221 5.69. The recomendations discussed here can be set out in a series of stages relating to different objectives. The implementation of such proposals would be best handled by a Tariff Commission. Stage 1. Estimate minimum required uniform basic tariff on revenue grounds; Stage 2. Identify luxury consumer goods imports that qualify for higher taxation on income distribution grounds; either a) increase the import tariff rate and offset this higher rate with an equivalent excise tax on domestic producers of these goods, or b) levy an excise tax applicable equally to both imports and domestic production; Stage 3. Identify industries that qualify for special protective tariffs, and estimate what rates are required; and 21/ A simple illustration is as follows; local company B uses tariff protection to sell 1,000 units of output at Mt 1,000 per unit to company A (the exporter). Company B's unit sales price is 252 above the CIF price which is Mt 800 per unit at current exchange rate. Company A receives a tax (or other) credit of 1000 units times Mt 200 price differential to offset this, or Mt 200,000. The tariff Commission would be responsible for approving such a credit. 22/ It should be noted that any single-nation proposals need to be assessed in the light of attempts to move towards regional tariff harmonization under the Preferential Trade Area, which the World Bank is currently supporting. - 77 - Stage 4. Calculate whether any credits to exporters will be required to offset higher input costs due to protection. 5.70. Table 5.4 shows a set of hypothetical rates of nominal protection and subsidy as discussed here in relation to the varying objectives a tariff system must meet. In term of the impact of tariff protection on production incentives what matters is the effective rather than the nominal rate of protection. Technical Appendix 5.1 shows approximately what the EPR will be for different activities under the proposed scheme of protection. 5.71. The objective of a protective system should be to aim for broadly equal EPR unless there are activities which the government has good reason to encourage or protect, and which should therefore receive above-average effective protection. In practice it is very difficult to plan a tariff structure around a series of target rates of EPR. The aim should be to start with a view of which activities should be encouraged particularly and to set nominal tariff rates which give approximately the desired result in terms of effective protection. The scheme set out in the table is an initial framework allowing for complications noted in Technical Appendix 4.1. Table 5.4: HOZANBIQUE - TARIFF PROTECTION AND OTHER TAXES AND SUBSIDIES Oblective Recommended Import Tariffs Goods and other Taxes and Subsidies Covered Revenue (a) 30S uniform import tariff All Distributional (a) 302 uniform import tariff Luxury Equity (b) 302 additional import Consumer goods tariff on luxury imports tc) equivalent domestic excise tax on local production Protection (a) 30Z uniform import tariff Products of (b) special additional infant industry protective rates of or protected either 102, 20X or 302 for their employment effect Export Promotion (a) import tariff drawbacks All Exports on imported inputs (b) credits on locally produced protected inputs used in export production _ 78 - S. Measures for Price Setting Conditionax Prices 5.72. By the end of 1989 direct price controls have been removed for all industrial goods, with the exception of soap, petroleum fuels, and petroleum derivatives. In addition beer and cigarettes are set by the government for tAx revenue purposes, and the prices of several basic industrial consumer goods, - flour, bread, sugar and cooking oil, - are controlled as these are sold through the ration system. Other industrial prices are conditional* in that enterprises can set their own prices on the basis of a cost plus formula, and revise these when costs rise subject only to ex-post referral to controlling Ministries. (See also chapter 3). Interviews revealed few indications that enterprises have had difficulty in getting their proposed price changes appiroved. This move to a system where some industrial prices are checked ex post, but not set by the government, is an important part of the policy of enterprise autonomy. Evidence that prices are now largely freely determined is shown in table 5.5. Table 5.5: MOZWBIQUE - INCREASES IN CONDITIONAL INDUSTRIAL PRICES 1986-88 at (Annual increase 2) Z Increase 1986-87 1987-88 1988-89 1986-89 b/ Wire 253 44 84 838 Iron bars 219 60 128 1063 Irrigation pipes 380 15 0 454 G.I. sheet 420 221 0 1571 Other pipes 380 277 0 1709 Ploughs 237 13 57 497 Radios 171 89 41 623 Dyes 191 380 41 1852 Varnishes 445 240 94 3500 Oxygen 429 111 38 1443 Electrodes 367 75 53 1150 Tires 552 71 53 1608 Salt 116 199 17 656 Copra oil 450 34 n.a. n.a. Soap 170 187 n.a. n.a. Exchange rate 1788 c/ Consumer price index 514 dI a/ Prices of first quarter in each year. b/ 1986 prices were controlled. s/ Calculated from MT4O per US$ to Mt7l5 per US$ (lst quarter 1989). d/ 1986 - 100. Sources M13 _ 79 - 5.73. The table shows that industrial goods prices have mostly lagged behind the change in the nominai exchange rate, but have risen well in excess of the consumer price index. This suggests that prices have been relatively free to rise in line with cost increases. | 5.74. It is . questionable whether it is advisable to continue the conditional pricng system in its present form for the majority of industrial goods, because a price policy based on costs of production pl4s a predetermined markup does not encourage producers to look to ways to lower costs, as a means of raising profits and expanding markets. 5.75. From the trade efficiency point of view it is desirable that the prices producers receive domestically bear a relationship to World prices. There is therefore a good case for a system which links the domestic prices of traded goods more closely to the world prices for similar goods. If in the future the demand for imports is regulated by tariffs and exchange rate policy, not licensing controls, this form of control will be redundant, since if domestic prices rise too far competition from imports will act as a restraining factor. However, assuming form of licensing continues in the medium-term, prices of competing goods in the world market should be used as a guide to domestic ex-factory prices. Revision of Conditional Price Formula 5.76. There is a good case for continuing the system whereby firms are required to submit unit cost of production estimates. This costing exercise is useful in its own right in forcing firms to work with accounting concepts. (Asset revaluation, discussed in chapter 3 and 6, is required if realistic cost figures are to be produced for existing enterprises). It is proposed however that the unit cost-plus-markup prices be accepted only if they do not exceed the cif price of competing imports by more than an agreed margin. Where the proposed domestic price cannot be set within the agreed margin then ex post it should not be approved, and a review would take place. 5.77. It is preferable to set ceilings on domestic prices at the level of the import parity price (cif vrice at the official exchange rate) plus a protective margin. In section _.,40 a 301 margin of domestic preference is used for illustrative purposes in assessing whether domestic prices are sufficiently high to justify the issue of licerses for competing imports. For consistency the same percentage would need to be used in setting price ceilings. Domestic prices ex-factory would then be allowed up to 302 above CIF prices of competing goods, but would have to be applied flexibly in order to protect enterprises in the short run. The margin would be subject to decrease in the event of further real devaluation of the Metical. 5.78. Over time with a gradual reduction in the protective margin price ceilings will come down to closer to import parity price levels. Further, if in the medium term tariffs rather than licencing can be used to regulate the demand fer imports there will be an automatic link between domestic and world prices. In such a case these price ceilings will become redundant, so that _ 80 ° this type of control would be a transitional measure prior to the use of exchange rate policy and tariffs to regulate import demand.. 5.79. This form of price settiing will be a fairly loose control over many enterprises because of quality differences between local and foreign goods. For example, local blankets, and toothpaste compete for a different segment of the market from imports, being of a lower quality and price, and for this reason one would normally expect prices of local goods to be lower than for imports. For commodities such as these this form of price ceiling would therefore not be effective. However there will be some comnodities for which this price ceiling will be effective even in the: short-rxn. The following table indicates a range of ratios of domestic to world prices for the firms surveyed. Table 5*6: MOZAMIQUE - COWAPATIVE OUTPUT PRICE RATIOS (LATE 1988) Domestic Cif Excess of ex-factory Import domestic price price (Mt) over import price (Z) Cement Mt 2067/ Mt 1500/ 38.0 50 kg bag 50 kg bag Wire, cable Mt 450/mt Mt 650/mt -44.0 Cartons Mt 530,000 Mt 730,000 -37.0 Cotton poplin Mt 11901m2 Mt 750/m2 37.0 Trevira Mt 30001m2 Mt 1200/m2 60.0 Pasta Mt 118/kg Mt :.20/kg -2.0 Laundry soap Mt 240/kg Mt 300/kg -25.0 sources enterprise Interviews. 5.80. The above data are illustrative in most cases and are based on late 1988 exchange rates. In some cases local products are apparently competitive with international prices, while others are well above. In the case of cement the domestic price is 38Z higher than CIF which is marginally greater than 'the illustrative margin of 302; 231 the sector Ministry would have to take a view on the acceptable margin, given long term prospects of cutting costs when the security situation improves. This applies to all enterprises for which the new pricing formtla does not allow full cos recovery. Generally the attempt to keep prices within a margin would stimulate efforts to lower costs or alternatively discourage investment in these activities. 231 However in this particular case the quoted price may have been lower than the long term world price. The DRC calculation in chapter 4 uses what is thought to be a more representative price. In reality it is often difficult to identify specific world prices, and it is advisable to assume a price range. ° 81 - Demand-basid Pricing 5.81. It is possible that the market will not sustain prJces based on either the cost or the margin formula in present conditions of recession caused by economic adjustment. In the case of food products there was reported to be adequate demand in 1989; however, as stated in chapter 3, durable goods demand has probably not been adequate. The rapidity of the changes that have taken place in the market make it difficult to predict levels of demand on the basis of experience over 1987 - 89 because the effects of the changes may not have worked themselves through. If demand will not sustain the set prices then the price formulae would become simply maxima, and would be redundant as long as the market is depressed, with prices effectively being freely determined. Summary of Recommendations on Conditional Prices 5.82. It is recommended that the conditional pricing system ex factory should be based on world prices (cif Maputo) plus a margin based on the difference between the official and Oshadow' exchange rate. For illustration this could be the cif price of a competing import plus 302. The target would need to be adjusted periodically in the light of exchange rate and relative price movements. 5.83. .11 prices calculated from the current cost-plus formula and below the ceiling would continue to be app-oved automatically. However, if cost-plus prices exceed the ceiling (e.g. CIF price plus xl) the government should review the price and determine whether a restriction is justified. F. Investment Incentives The Current System 5.84. The 1984 and 1987 laws set out a generous system of tax exemptions and other conce,;sions for new investment by the foreign and domestic private sector. The main forms of incentive are tax holidays for periods of two up to a maximum of ten years. various tax credits relating to depreciation allowances, and training expenditure, and exemptions from import duties on capital goods and other materials. 5.85. The level of investment is determined by a number of factors, of which incentives are only one. In most countries the level of political and economic stability is likely to be the dominant influence. However, there is some evidence, discussed in chapters 3, that there has been an increased response from foreign investors to the 1984 law. As yet it is too soon to judge the impact of tLe 1987 law on the domestic private sector, or the 1988 law on debt equity swaps (discussed in chapter 6). - 82 - 5.86. Prior to the December 1988 tax changes relatively high nominal rates of tax on profits and income, (50Z on profits and 552 on distributed dividends), were combined with significant exemptions for new investment. Existing producers not able to undertake new investment were subject to these relatively high rates without exemptions. The level of the basic rates of tasatio. was identified as a significant disincentive by private industrialists during the industrial conferences of 1987 and 1988 (see chapter 3) and from those interviewed, who drew attention to the lack of exemptions for existing producers. There was a case therefore for lowering the basic rate of profits tax, whilst reducing and rationalizing the system of exemptions. A reduction in exemptions could increase Government revenue, andlor allow.a lower basic rate of profits tax for existing producers. In this way producers not. currently considering major investment could receive an incentive from the tax system. 5.87. The removal of the dividends tax of 55? in December 1988, and the restriction of profits tax to a single rate of 50? are major steps in this direction. However, this lower overall level of taxation strengthens the case for rationalizing and limiting exemptions to the basic tax on profits. Furthermore, when tax revenue consideration permit, the Government could consider lowering the basic profits tax rate to 40?. 5.88. The rate of profits tax is still relatively high by the standards of other countries where average nominal rates are in the range 40? to 45?. 24! When th% objective is td stimulate private sector investment a lowering of the basic rate say to 402, combined with the removal of the dividends tax, could have an important incentive effect, particularly for existing producers. Rationalization of Tax Ezemptions 5.89. In order to regularize tax liabilities there is an initial need for annual revaluation of company assets, so that depreciation allowances can approximate more closely the economic use of assets. A law to that effect has been introduced (see chapter 6). 5.90. There is a case for limiting the period of tax holiday in order a) to increase tax revenue and finance a lower rate of profits tax, and also b) to discourage investment by foreign firms who recover their capital over the initial tax holiday period and then move their operations to another country. Tax holidays can also have the effect of transferring revenue to foreign governments who would otherwise exempt profits from tax if they have already been taxed abroad. (In other words the tax holiday for the investor may be offset by tax in his home country so that the net incentive is weakened). This implies the need for a limit to the concessions offered through this mechanism. 5.91. The aim must be to encourage investment that will provide stable employment and tax revenue over a relatively long time-period, and which does 24! Anderson 1987; op cit p.13. _ 83 - not move between countries once the period of tax exemption comes to an end. (An IMF proposal is for a maximum figure for the tax credit allowed under the tax holiday arrangements, defined as a percentage of the original capital investment of 150Z over a maximum of 5 years). Despite the general case for limiting tax holidays it must however also be recognized that government discretion over the tax concessions offered to foreign firms is constrained by the competition from other countries wishing to attract foreign investment. In the short term it is advisable to continue to allow flexibility in negotiations between the government and foreign firms. The potential economic return to the country and the rate of profit expected by the foreign investor will vary between investments, and the government will nee.d to be able to offer an incentive package that meets the circumstances of the particular case, if this is practicable. 5.92. An additional factor that must be borne ir. mind Is that government tax revenue, along with higher employment, are likely to be the main direct national benefits from foreign investment and that high tax exemptions will reduce the government revenue gains from such investment. Therefore tax holidays of 5 years or more should be treated as very exceptional, with 3 or 4 years taken as a normal period. 5.93. In the case of domertic investment the government has more control over tax concessions, since the issue of competing incentives abroad is less relevant. The present incentive legislation relating to domestic private investment is generous allowing tax holiday periods of 3 to 10 yeiars during which no profits tax is paid. In addition double depreciation allowances for between 5 to 10 years are also available. These incentives appear excessive, and it is recommended that the period of tax holiday. should be reduced. The purpose of the double depreciation allowance is to reduce or eliminate profits tax liability in the initial years of an enterprise. In some countries a contradiction has arisen whereby tax holiday incentives and depreciation incentives are mutually exclusive, so that the tax incentive is partly redundant. However, the regulations in Mozambique prevent this. 5.94. Several aspects of customs tariff exemptions under the existing incentive legislation could also be modified. Capital goods imports for new investment are allowed duty-free for foreign investors, and for domestic investors where their own resources are involved. Under the domestic investment law, where a domestic investor does not use its own foreign erchange, 50Z to 60Z of the duties eligible are exempt. This incentive has the effect of encouraging the use of more capital relative to labor, depending on the type of industry and the initial absolute levels of capital costs and money wages. It is not justifiable to promote investment in a way that encourages greater use if capital relative to labor inputs. It is recommended therefore that the incentive laws should be modified by gradually withdrawing tariff exemptions on capital goods. This means that, should a tariff reform programme such as that outlined above be adopted, capital goods imports would be eventually subject to the standard rate of import duty. The reduction in the basic zate of profits tax would partially _ 84 - compensate for the loss of exemption in terms of Governmeut revenue. 251 5.95. There is also a case for removal of tariff exemptiou on raw material irports, which it is understood are allowed duty-free under the domestic investment law, and for exporters under the foreign investment law. Such exemptions can be justified in the case of exporters, for reasons outlined in section 5.66, but for other producers already receiving protection from the tariff system it is desirable not to exempt raw materials, since the exemption would raise significantly the effective protection these producers receive. This would mean that, under the proposed tariff reform, material imports would be subject to the standard uniform rate of duty - unless they are subject to a special protective rate. In this way the revenue base of the customs tariff will be less eroded than under the existing set of exemptions. Employment Tax Credit 5.96. A possible long term option is to use a labor subsidy as the main form of tax incentive offered to producers for new investment. This could operate by granting firms a tax credit, which could be deductible from their Profits Tax liability, or through other means. The tax credit could be directly proportional to the number of workers employed as a result of a new investment, subject to safeguards against overmanning. In this respect it would parallel the existing allowance against costs of training personnel. A total annual figure for the credit would be obtained by multiplying the number of woicers employed in the firm as a result of the investment by a value per worker (related to the minimum wage). This value per worker would be given as a percentage of the minimum wage (normally in excess of 1002), although its size would depend on how generous the goverrment-wishes to be in reducing tax liability. For example, suppose a firm employs 100 workere as a result of a new investment and the Government allows a credit of 1.5 times the minimum wage rate, which is set at MT 12,000 per month. Therefore the annual credit would be (MT12 x 12,000 x 1.5 x 100) or MT21.6 billion. 5.98. It is necessary in estimating the tax credit to use the minimum wage as a reference point, to avoid firms artificially inflating their wages bill to qualify for a higher tax credit. However enforcement would still require an accurate check on numbers employed per firm. The aim of granting tax concessions in this way rather than through a simple tax holiday, is that this incentive favors firms whose wage costs are high relative to both profits and value-added. Thus the more labor-intensive producers would benefit. 5.99. If the value per worker used in calculating the credit is high relative to the minimum wage, this tax credit scheme could be used to replace 25/ It would not be advisable to withdraw the exemption completely in the short run as this could have adverse effects on the recovery of investment. A phased program could be considered. - 85 _ simple tax holidays. Therefore instead of granting a tax holiday for 2 years, a firm could be granted 2 years tax credit calculated in this way. Whether firms would gain from the switch to this scheme would depend both upon their labor-intensity in production and the level of and period for which the tax credit would apply. The illustration in Technical Appendix 5.2 shows how this tax credit scheme could work for a hypothetical project. 5.100. It is not recommended that a tax credit of this type be introduced at present, but that the government may wish to consider it in the long run as an alternative to holidays, since it is a relatively simple means of encouraging employment generating activities. The Effects of the Turnover Tax (Imposto Circulacao) 5.101. The turnover tax, which is levied each time goods change hands in the process of production and distribution, results in a 'cascading' effect. This is a well-known form of distortion, since it penalizes specialization and encourage' producers to integrate backwards as a means of reducing the taxation imposed when inputs are purchased from outside suppliers. Because the tax accumulates at each stage of production it may also encourage imports of finished goods and therefore discourage local production using imported components. Similarly, without appropriate exemptions it could tend to encourage exports of raw or semi-finished materials rather than finished products in activities which are not vertically integrated. In either case there may be a disincentive to domestic production which amounts to negative protection. The effects of the tax can be seen from an example. Table 5.7t MOZAMBIQUE - ILLUSTRATION OF EFFECT OF 10 TURNOVER TAX MtOOO per unit Case 1 Case 2 without tax with tax without tax with tax Supplier - input B 10 li 10 11 -.input C 20 22 20 22 Labor 10 10 10 10 Tax on A's sales - 4.8 - 5.2 A's profit after tax 8 0.2 8 8.6 A's sales price 48 48 48 56.8 Total turnover tax - 7.8 - 8.2 5.102. In case 1 'with tax' it is assumed that enterprise A cannot raise its price to pass on the tax, while A's suppliers, B and C, do pass on total tax of Mt3,*000 per unit to A; A pays Mt4,800 tax on its own turnover. The result is a total turnover tax of Mt 7,800 per unit which almost'eliminates the profits which A would earn if there were no tax. In case 2, A passes on the tax, resulting in a higher price. Case 1 is likely to be the more _ 86 - realistic situation in Mozambique because local manufacturers are mainly *price-takers., facing set import prices. Competing imported products would only have to face the final stage turnover tax and would therefore have an advantage. For example with an import market price of Mt4O,000 (equal to A's production costs) the imported product would only incur turnover tax of about Mt4,400 per unit (about 10Z of the sales price 44,400), giving the importer a price advantage of Mt3,600 per unit, equal to about 72 of A's sales price. 5.103. It has been suggested that the turnover tax should be replaced in time by a value-added tax (VAT). 26/ However, it is also rccepted that the capacity of the tax system to absorb major changes is limited. Introduction of a VAT system would create significant administrative difficulties. and such a reform would on these grounds be justifiably deferred. It could be reexamined when any further major changes are made to the tariff structure, since such changes will also necessitate some alteration in domestic taxes. In the interim, methods of offsetting the potential import bias of the existing system on specific products should be considered, possibly using a system of credits. Access by Small Enterprises to Tax Incentives 5.104. Discussions with the MIE and the firms themselves suggest that provincially registered firms (below about 50 workers) are not utilizing the tax incentive legislation. This is a common problem with incentives ^%f this type - which require firms to be fully integrated into the tax accounting system. Often, where small-scale firms are not fully covered by the tax network, they will prefer to avoid applying for incentives but rather minimize their tax bill by under or non-recording. The tax system may not therefore be available as a mechanism for supplying incentives to smaller enterprises. Support for such enterprises have to rely more on institutional assistance from bodies like IDIL and the Provincial Directorates, (see chapter 6), and over the longer term on general improvements in macroeconomic conditions. Effects of Proposed Tax Changes 5.105. When the changes considered above are combined with the recent abolition of dividends tax, they will have the effect of altering the time profile of a firm's tax payments, with more tax paid in the short-run, due to lower exemptions, but less in longer-term due to the reduced tax on profits. This would be accentuated with further reduction in profits tax, to e.g. 402. This will be a change more favorable to producers who operate with a longer time perspective and who are willing to commit resources to productive uses for a relatively long period. It is precisely this type of private investor the government should be encouraging. 5.106. From a public revenue point of view there will be an advantage in this approach since the government will gain tax receipts in the short-term, 26/ Source: DMF (1988) op cit p. 35; - 87 - with the lower receipts from the lower rate of taxation spread over the medium and long-run. This means that even if the changes proposed created exactly the same total tax revenue as under the present system, the timing of the tax payments will be such that the discounted value of the tctal tax flow will be higher than under the present system. Since government income can be used in various ways it has an opportunity cost, so that it is appropriate to discount tax flows to a present worth. Technical Appendix 5.2 illustrates this point using the example of a hypothetical project. Under the set of incentives proposed here total tax revenue is only slightly higher than under incentives similar to the existing system. However the present value of the tax receipts (discounted at 8X) is considerably higher under the new set of proposals. Summary of Recommendations on Taxation 5.107. The proposals for a reduced and rationalized set of exemptions. and the other considerations regarding the tax system are as follows: (a) there is a case for reducing the normal period of tax holiday on profits tax for foreign investors to 3 to 4 years, with more than 5 years only granted in exceptional circumstances. However the government could keep flexibility in its negotiating position by retaining the existing legislation which sets out a wide range of tax holiday periods; (b) there is a case for reducing the maximum period of tax holiday on profits tax for domestic investors to the lower end of the present range of 3 to 10 years; (c) the Government should consider the case for gradually withdrawing im4ort tariff exemptions on capital goods imports and raw material imports for non-exporting fimss and levying tariffs at a standard rate of the type recommended in section 5.4; (d) because of its discriminatory effects on local products vis-a- vis imports, the possibility of replacing the turnover tax (imposto circulacao) should be reexamined when tax reforms are contemplated, with possible interim use of tax credit for particular products; and (e) research should be carried out to determine whether small-scale enterprises face difficulties resulting from lack of access to fiscal incentives, and, if so, how they can be incorporated into the system. - 88 - VI. THE BANKaNG SECTOR AND TEE PINANCING OF INDUSTRIAL REHABILITATION A. The Role of the Flnancial Sector 6.01. The strategy for the reform of the financial system in Mozambique requires an understanding o' 'a) the role of financial institutions in a market economy, and (b) the process of transition from a financial system which is dictated by a central credit and production plan to one which functions within a competitive market. This chapter addresses some of these issues as they relate to the finance of industrial rehabilitation. 6.02. Generally, the financial sector in a market economy acts as the intermediary for the mobilization of savings and the allocation of investment, both domestic and foreign. It also facilitates trade and specialization in production, and therefore improves the efficiency of resource use. The interest rate, which determines the yield on savings and the cost of investment funds, is central to the system, as it effects the rate of savings and investment, and therefore, ceteris paribus, the rate of economic growth. Economies with distorted (repressed) interest rates have been shown empirically to grow more slowly than those with market-determined rates, because negative real interest rates discourage domestic financial savings, and encourage hoarding and capital flight, while at the same time they permit relatively unprofitable investments. 1/ 6.03. Higher interest rates attract surpluses into the financial system, bringing about financial deepening. Industrial countries tend to have "deeper* financial sectors than developing countries. 2/ In addition there is an inverse relation between the level of net financial assets as a percentage of GDP and the level of external debt as a percentage of GDP. This is caused inter alia by the fact that countrics which have undeveloped domestic financial systems have a higher propensity to borrow abroad. Research in fourteen middle-income developing countries showed that the financial sector was responsible for financing about 45Z of corporate investment over recent years. using principally funds deposited in banks and other institutions by the household sector. 3/ 6.04. An efficient financial system requires rational interest rates and competitive institutions. It also needs to develop an environment in which lending risks and transaction costs are minimized. This requires the development of infrastructure, including a range of financial instruments, sources of financial information, accounting and auditing of adequate 1/ This is discussed in more detailed in World Development Report, The World Bank. 1989. 21 Financial deepening is measured by the growth of money supply M2 (currency and bank current and term deposits) as a proportion of GDP. 31 Source: World Development Report, 1989, Table 2.2. _ 89 - quality. and, finally, a legal and regulatory system which inter alia ensures enforceability of contracts. The range of financial instruments would include different types of funding, such as equity and quasi-equity, as well as loan finance, public and corporate bonds etc. Equity capital shifts financing risks from the investor to the saver, which is important in an infant industry situation. A set of regulatory agencies is needed to control the financial system, in order to avoid misuse of savers funds and ensure acceptable lending practice. A number of initiatives have been taken recently in Mozambique which signal the beginning of the revival of an effective financial sector. S. The Operation of BankinR Institutions in Mozawbjgue 6.05. There are currently three banks in Mozambique:- the Bank Of Mozpbhique (BM), the Banco Popular de Desenvolvimento (BPD), and the Banco Standard Totta (BST) which is private. The other main financial institution is the insurance company, Empresa Mocambicana de Seguros. In 1988 the Caixa de Credito Agrario e Desenvolvimento Rural (CCADR) was established, as an affiliate of BPD, which is a fund channelling subsidized credit to the rural sector. BM is by far the largest bank. acting as central bank, issuing bank. state bank and foreign trade bank; it has also functioned as the main commercial bank. It is responisible for proposing credit, interest rate and exchange rate policie' to the Government and has had monopoly control of foreign exchange transactions until recently. The BM has a total of 10 provincial 'filiales", 17 district branch offices, and 13 deposit offices. 6.06. The BPD is 1OOZ state owned, formed at independence from 3 nationalized banks. It has a total of 10 'filiales' and 138 branches and collection posts. BPD has concentrated on short term commercial agricultural finance, but under its statutes it is empowered to lend for medium to long term investments; it can alsc invest in equity, take up commercial paper, and promote new ventures. It has equity holdings ia a number of textile enterprises 41 and in some srsll firms and in 1988 established a share capital participation department which has taken over responsibility for its equity holdings. BST is privately owned. 5/ It has reduced its branch network drastically since independence because of lack of business, f.om 45 to 6, and n recent years it has been lending for short term purposes only. 6.07. Over the period 1980-86 the banking system acted as a channel for funding according to the financing needs of the central plan, bypassing normal bank lending criteria. Thus, while GDP fell in real terms over this period, domestic credit expanded rapidly, and serious arrears were built up. Interest rates were not used for rationing of credit and were controlled at rates well below the rate of inflation, in the range 4 to 8 for enterprises, depending on the borrowing sector. Bank deposits were very largely demand deposits which officially earned zero to 22 interest. The situation was not conducive to mobilization of real savings and would normally have led to 41 These are Texmoque, Texmanta, Texlom, and Riopele. 5/ The shareholders are Standard Chartered UK (302), Totta E Acores, Portugal (302) and a group of other holders. - 9o - excess demand for investment funds; however the nonavailability of foreign exchange for import of investment goods effectively brcught long term investment to a halt and liquid reserves were accumulated which in the case of the two smaller banks, BST and BPD, increased steadily in nominal terms to exceed 302 of total liabilities by 1987. These liquid resources were partly used to finance BM's lending; thus Central Bank control of the financial system, e.g. through a rediscounting facility, was redundant. 6.08. Over 1981-86 large lending arrears were built up, especial;y by BPD on loans to the state farm sector and by all banks lending to finance the operating losses of enterprises subject to price and production co trol. Total arrears of the banking system were well over 50Z of total loans in 1986. Loans to enterprisas to cover operating losses effectively became subsidies, and normal loan security criteria and repayment obligations were generally bypassed. Banks were able to maintain financial stability because of high levels of liquidity. (Nominal deposits expanded despite negative real interest rates, reflecting inflation on parallel markets, although in real terms, bank deposits fell.) 61 High yield margins over the average cost of deposited funds assisted in maintaining eari'ings. However, even with nonperforming loans included on balance sheets at full value, the capitalization ratio of BPD and BST is not adequate to provide minimal security to depositors in the event of actual loan default. The reduced value of portfolios after adjusting for nonperforming loans and the inadequate loss provisions mean that the banks are technically insolvent. At present BM's loan portfolio is highly concentrated, with 30 companies accounting for 502 of outstanding borrowings, and the Railways alone accounting for 142. Lack of portfolio diversification involves relatively high default risk. 6.09. The result of foreign exchange shortage, distortion of interest rates and the other problems of the banking system is that industry has been u-nable to use the banks to finance longer term borrowing for investment, even where opportunities have arisen. Thus potentially profitable investment opportunities may have been missed while the limited funds available were channelled to financially and economically unsound investments, mainly in the staZe sector, or left as unproductive cash holdings. Lack of availability of loan finance may also have discouraged enterprise owners from investing equity, since, especially in the case of potentially relatively profitable ventures, the "leverage' from using loan funds boosts the return to equity and is sn inducement to investors. 6.10. Over the period 1980-86 Lending was concentrated in BM, which proviled 75 to 80Z of total bank credit, in addition to its functions as Central Bank. The concentration of lending contrasted with that of the private sector bank, BST, which reduced its share of lending to 32 by 1986, along with its "ranch networ!e., concentrating exclusively on short term lending. Lending by commercial banks to BM grew steadily over 1981-86 and in 1987 amounted to about 302 of BM's total deposits. The remaind'r of its deposits were from the corporate and personal sector. BPD accounted for the 6/ The monetary aggregate M2 (currency, demand deposits and time deposits) as a proportion of GDP rode from 45Z in 1980 to 79Z in 1983 and then fell back to 662 in 1985 and 67Z in 1986. - 91 - remaining :72 of lending in 1986, which went largely to the state farm sector. The dual role of the BM was not conducive to the development of its regulatory and policy making respotsibility. The following figure shows the distribution of bank credit to non Government borrowers over 1980-89. FIGURE 6.1 0.0' IM I W .,m 0.4 0.3 0. 1U7 193 Source: IMF/ Bank of Mozambique. 1989 shares are for 6 months only. 6.11. The figure, shows bank lending shares over time, including both domestic credit an*i external funds onlent by the banks. Since 1987 a significant change in shares has occurred, largely as a result of the establisement of the CCADR. Starting up in 1988, the CCADR accounted for 24% of new lending by the first half of 1989. BM's share of new lending fell to 511, while BPD's share levelled out at around 21S. BST, which showed signs of a resurgence in 1988, fell back to 3.42 of new lending in 1989. This representz a decentralization of lending from BM, but the state sector remains dominant, expanding lending with subsidized credit through CCADR, largely to agriculture. 6.12. Industry and agriculture have accounted for the bulk of bank lending, around 702 over the period 1984-89. l/ Since 1987 industry's share has increased, reaching 352 of new lending in the first half of 1989. BM has accounted for the majority (up to 902) of lending to industry, agro-industry and coinerce. BPD lending was concentrated in agricuilture until 1987 but it has since increased its share of industrial lending. BST lending was largely to the urban trade se^tor. l/ These two subsectors were reclassified in 1986 and therefore their separate shares cannot be compared directly with previous years. _ 92 - 6.13. The implications nf the changes in share for the financial system are as yet unclear. The development of the sector cannot be assisted through subsidized credit. However CCADR is not strictly a bank, but rather an institution currently used for channelling aid funds. As such it has no obligations to depositors and is dependent on Government financing policy. In 1990 budgetary funding is expected to be restricted. The more important indicator of financial diversification and strengthening will be the level of private sector activity. One positive indication of the revival of lending has been a shift to long term loans by all banks. from 122 of all loans in 1987 to 242 in 1989. BST, which has been almost exclusively a short-term lender, has raised its long-term lending from 32 to 142 of new loans. C. The Reform of the Banking Sector 1 6.14. The Economic Rehebilitation Program has resulted in a number of actions which have affected *he financial sector, related firstly to stabilization and secondly to structural reforms. The principal economic stabilization measure has been the setting of credit ceilings for the banks, designed to ensure that monetary growth is consistent with fiscal and external payments targets. Economic Stabilization; Credit Ceilings 6.15. Credit ceilings were imposed under the ERP in order to assist in reducing the fiscal deficit and restraining price inflation following the devaluation of the currency. The utilization .f ceilings on net new lending has been as follows: Table 6.1s MOZAMBIQUE - BANKING SYSTEK; DOMESTIC CREDIT CEILINGS AND UTILIZATION a/ (Incremental Credit - Current Prices) 1987 1988 1989 1990 Overall ceiling 40.0 74.0 68.0 78.0 Credit to Government 25.0 20.0 15.0 3.0 Credit to Economy 15.0 54.0 53.0 75.0 Actual domestic credit 33.5 59.9 68.0 to Government (actual) 14.4 3.2 9.0 to Economy (actual) 19.1 56.7 59.0 Total utilization 2 83.72 80.92 100.0S aI These figures exclude onlent external funds. Source: Bank of Mozambique. 6.16. The domestic credit ceilings and utilization levels represent large real reductions, particularly in the case of lending to the - 93 - 6.16. The domestic credit ceilings and utilization levels represent large real reductions, pprticularly in the case of lending to the Government. 8I New credit to the Government declined from a target of 622 of total lending in 1987 (actual 432) to a target of 22Z (actual 13Z) in 1989. Reduction in Government borrowing from the banking system is an important indicator of monetary stabilization, and also means reduced crowding cut of the private sector. In 1987 and 1988 new domestic credit to the economy was below its :eilings, and declined in real terms in comparison i rith earlier years. 9/ The fact that domestic lending apparently kept within a declining real target may imply a deficiency in overall demand for goods which has implications for industrial recovery, discussed in chapter 3. However, the real value of liquid assets within BPD and BST was maintained in 1988, which does not signify lack of purchasing power. 10/ (However, there may have been absorption of liquid assets previously held outside the banking system.) Structural Reforms in the Financial System 6.17. The principal longer term reforms in the financial sector which have been initiated under the ERP have included the following. (a) Interest Rates. Rates have been increased, with a simplification of the rate structure. The objective is to attain positive real rates in 1990. With the decline in inflation this objective appears to be on course. Time deposit rates are in the range 12 to 20S depending on deposit period with 32 theoretically payable on demand deposits. As of the end of 1988 the range for term lending has been adjusted to 22Z to 26Z for productive sectors, excluding transport, distribution and hotels which carry an average 33Z. However, further realignment is probably appropriate, in the case of the Central Bank rediscount rate (currently 10 to 12Z) and the rate at which the Government borrows from the banking system to finance the budget (currently 8Z). These issues are under discussion; 8/ Total credit as a proportion of GDP fell from 1012 in 1986 to 472 in 1988, reflecting very large falls in credit both to the economy and to Government. From 1987 to 1988 credit to Government as a proportion of GDP fell further while credit to the economy levelled out. 9/ The ceilings applied to lending net of repayments. However, the Government has been reimbursing BM for enterprise debt obligations which it has assumed. This repayment* has therefore permitted higher levels of gross lending. Taking this into account the utilization of credit was higher than planned. 10/ Nominal value of ligujid assets kept pace with price inflation, rising by 502, to 4O) of the total assets of BPD and BST. - 94 _ (b) Raumlato.Iy Control. A new accounting system has been introduced iito the banks and the DM has instituted a new organization structure dei.gned intet alia to permit separation of its commercial bankng' from its rentral banking functions. This would allow overall policy, supervision and regulation of lending to be divorced from lending operations, and would assist ia the establishment cf a regulatory system for the commercial banks. (c) Developnent of Financial Markets and Instruments. Experimental Ussues of tzeauury bonds were made in 1988 and 1989. The aim of 'his was to reduce the excess liquidity in the banking system, :ir to'begin to develop new instruments for public finance. The ain purchasers were enterprises and households looking for a superior yield to that available on time deposits. 83Z of the i88 issue was handled by BPD, which is starting to develop a financial services capability. (This is discussed in chapter 6). The use of bonds is an important step in the rationalization of the interest rate structure because it will help to equalize private and public sector borrowing rates A further indicator of change is the reintroduction of discounting of trade bills and consumer financing, and the increase in the proportion of long- term lending; and (d) Enterprise Debt Restructuring and Bank Recapitalization. Lending arrears are in the process of being cleared. In 1988 about Mt 8.8 bn (or 22Z) of Mt 40.6 bn of eligible lending arrears of the Bank of Mozambique were assumed by the Government, and a further Nt 10.0 bn were under discussion. About 10 public sector or intervened firms were included in the first transfer. This debt may be converted into long term instruments at nominal interest rate. The remaining arrears will require a renegotiation between the banks and companies (intervened and private) concerned. Arrears with individual companies have also been rescheduled by BM. However currently debt transfer has been restricted to state and intervened enterprises, rather than the private sector. Decontrol of prices has permitted the possible recovery of nonperforming loans, and further financing of enterprise losses is now a fiscal matter. In 1988 this applied principally to the railways. Glass and cement industries were also financed through the budget. These initiatives are critical to the effectiveness of the banking system, and to the restoration of financial discipline. ill 6.18. Future development of the financial sector will require continued longer term structural reform, and stabilization through recapitalization, and portfolio restructuring through transfer to the budget, improved collection, rescheduling, or financing of write-offs. According to the 1X/ Interviews at the end of 1988 with branch managers and other officials involved in lending suggested that debt service discipline on credit disbursed since January of 1987 has significantly improved. - 95 - assets at 1988 values. This requirement would be reduced to the extent that (a) the Government takes over and finances additional debt, (b) overdue loan recovery exceeds expectations, and tc) loans are rescheduled. Clearly, financing on this scale would have to be phased over a number of years. Along with reestablishment of sound banking, options for expansion of the sector and development of new institutions would then be considered. 6.19. The future financial position of the banks will have to be safeguarded through the imposition of profitability criteria for lending, and also through the imposition of a regulatory system which the Government is currently considering. The regulatory system will have to lay down performance criteria for (a) capitalization and leverage (equity in relation to assets), portfolio quality (arrears rates), and liquidity (cash and liquid assets in relation to total liabilities); there may also be criteria for the maturity structure of loans and deposits. D. The Demand for Industrial Credit Real Domestic Lending 6.20. New domestic lending to the economy, in nominal and real terms, over 1985-1988 was as follows: Table 6.2: MOZAMBIQUE - NET DOMESTIC LENDING TO THE ECONOMY a/ Rnd 1985 1986 1987 1988 (Mt billiona) (Current prices) Total domestic lending 82.9 99.2 113.9 162.7 Increase in lending 8.2 16.3 14.7 48.8 (Const prices) Consumer price index 100.0 113.2 297.7 447.1 Total leading 82.9 87.6 40.4 38.5 Increase in real lending. 8.2 4.7 neg neg at Excludes external funds onlent. Sources IMP and study estimates. 6.21. There was a reduction in the real value of total domestic lending, with negative real net increases in both 1987 and 1988 due to the imposition of the credit ceilings. Thus historical data on domestic lending gives no guide to future investment demand. This is both because of the credit ceilings and because lending has been short term. Indications are that credit demand reached ceiling in 1989, but this is difficult to interpret. _ 96 - Industrial Credit Survey 6.22. To gauge the requirement for investment lending a credit survey As carried out to obtain some indicative information about credit demand. e:r.imated financing requirements of a number of firms were highly tentative, and it could not be ascertained a priori whether the enterprises responding would qualify for loans on profitability or economic efficitncy grounds. Furthermore it is not clear to what extent enterprises require credit as such rather than simply access to foreign exchange (financed through existing cash reserves). However, in view of the decline in the real value of cash balances over the past 2 years it is likely that a substantial proportion would represent a demand for credit. The data may be therefore regarded as a starting point. 6.Z3. Tne total estimated requirement for 119 enterprises was US$61 million, excluding a single proposal of US$10 million for one enterprise. 76 responding enterprises were at *national" level (roughly above 50 workers), and 44 at *local" level (below 50 workers) of which all were in the Maputo area. Total requirement at larger scale (excluding the largest proposal), was US$55 million, and at small scale was US$6.0 million. Average loan size for the larger scale group was US$733,000, of which US$401,000 was for fixed and US$332,000 for working capital. For the smaller scale group average financing requirement was US$136,000, of which US$83,000 was for fixed and US$53,000 for working capi--l. 6.24. The following table summarizes the credit survey results. An estimate of capital requirements had to be made for 30Z of the companies responding on the basis of averages for working capital and fixed capital per sub-sector. Companies are grouped by sub-sector and by the government organization which is responsible. 12/ Of the responses 39 were received from the Executive Council of Mapuato City, which is responsible for industries with up to 50 employees. For this reason these industries are denominated as 'small scale industries*. 6.25. The number of industries responding are divided into two columns: Column (1) shows total industries responding, and column (2) shows industries stating requirements but providing inadequate cost information. Credit demand estimated from average working fixed capital requirements for column (1). 12/ The Government organizations which sent responses are: Secretaria do Estado Para a Industria Ligeira; Hinisterio de Industria e Energia; Ministerio de Construcao e Aguas; Ministerio de Agricultura; Ministerio de Informacao; Conselho Executivo da Cidade de Maputo. 97 - Table 6 3: MOZAEIQU - SURVEY RESULTS I SMWARY DWID FOR INWUSTRIAL CREIT (1988) (US$000) (1)(2) (1) (2) Sector and Govt .W.C.o F.C.!/ TOTAL Avg Avg EST. EST. EST ReupAndont. Organization TOT TOT W.C. F.C. W.C. f.C. TOT (NOQ) 14 6 food/tobacco 1184 1672 2706 140 200 640 1200 2040 6 metel products 6290 1613 7608 1260 800 - - --- 9 2 metl product. 8946 126 4761 600 160 1000 860 1860 6 off cI *uppl too 2S69 1163 8n7 480 190 -- 6 8 sto condition 8650 1069 4919 60 160 -- ---- ---- 3 1 petroleum - 802 802 --- 150 150 150 8 6 printtng S60 720 1800 190 140 760 650 1820 12 4 con.truction 1411 1650 8061 800 240 1200 960 2160 4 1 agriculture -- 8656 8565 - 1200 --- 1200 1200 6 1 other 280 56114 8419 60 1120 650 1120 1680 a other - 466 4685 200* ---- -- - --- 76 20 TMTit 221BC 28183 462B6 6 S6S0 9910 Smal IIndust lee b 11 4 food 70 726 709 10 100 40 400 440 10 6 metal product. 486 286 721 110 70 6S0 420 1080 8 8 cosmeticc 501 897 606 100 60 8OO 240 540 5 2 clothing 150 489 589 50 150 100 8O0 400 2 1 furniture - 20 20 -- 20 -- 20 20 5 eho*e/tannery 78 68 161 15 15 - 3 2 othere 28 67 95 SO 70 6O 140 200 44 18 12683 2017 B800 1160 1380 2540 120 88 Overall tot.l 23898 25160 48638 5620 8930 12460 No"T: For local coate exchane rete at tim_ of survey Mt 600 W1. a/ W.C. a working capital. F.C. u fixed capital. b/ Solos 60 worker . 6.26. The above table excludes a $10 million proposal from Cia. Industrial de Hatola. On the above estimated basis total credit demand would then be: Sooo Companies stating costs of projects: 48,568 Estimates for companies not stating costs: 12,450 Compania Industrial de Matola: 10,000 Total 71,018 In addition to the $71 million compiled above further requirements were noted from companies interviewed outside the survey, of some $30 million. 6.27. Direct extrapolation of the sample of 75 national level industries for the sector as a whole (580 enterprises) cannot be done due to the degree of uncertainty of the sample estimate. If it is assumed that 25Z of the sample estimate for larger enterprises of about US$55 million (excluding Matola) constituted "realistic" credit demand, then extrapolation would lead _ 98 - to a total requirement of about US$110 million for fixed and working capital. Given the time period needed to implement major replacements, or new investments, the phasing of such a requirement would probably be realistically over a 3 year period, or US$30 to $40 million per annum. ExtrapolatLag from the 44 olocal enterprises (US$6 million requirement) to the national estimate of about 600 enterprises in the size range 10 to 50 workers (table 2.5) is even more haz.rdoue since the sample i8 considerably less representative. A lower bound requirement of US$20 million is a conservative starting point for this category. The combined demand from bankable and efficient enterprises in the two categories would be, very conservatively, US$130 million. Some support for this as a minimum estimate comes from the alternative projection of $113 million for 15 larger enterprises by Arthur D. Little. 13/ 3. EnterDrLse Restructurings Revaluation of Assets Objectives 6.23. To assist the transition two areas of Government action in enterprise rehabilitation are in process: renegotiation of debt granted pre- 1987 (discussed above) and revaluation of assets. The need for adjustment of asset values is best seen in three areas vital to the industrial sector: (i) Long Term Credit and Investment: drastically undervalued assets make credit analysis LmpossLble as companies appear unable to guarantee debt with equity because of the dlstorted relative size of the debt to total asaets of the company. (ii) Taxation: undervalued assets generate negligible depreciation allowances which in turn causes over-taxation and decapitalization of companies. this is particularly true in Mozambique given its high taxation of profits. (iii) PrivatLzation: Since 1986, as stated in chapter 3, the Government has sold a number of enterprises to local entrepreneurs. Without revaluation a strong possibility exists that enterprises would be sold at unrealistically low prices. To ensure that the price paid is fair value revaluation of assets will be an important indicator. 6.29. Glven the age of the industrial assets in Mozambique and the lack of investaent since the late 1970s it is also possible that a process of revaluation could overvalue the assets of these companies. However, for the purpose of bringlng greater coherence to values in the economy limited overvaluation is a relatively small distortion compared to the pres1.it situation. An illustration is given below. 13/ For the Industrial Enterprise Restructuring Project. - 99 - Asset Revaluation; A Case Study 6.30. For illustrative purposes the case of Mabor, a tire manufacturer, is used. This case demonstrates typical problems of valuation. Habor management estimated that initial investment was $30 million in the period 1975-79. Replacement cost according to technicians from General Tire, which oversees production in the plant would currently be $50 to 70 million. Despite its age, the plant and machinery is in good condition and able to produ^e its full rated capacity. 6.31. The following is a comparative analysis of the balance sheet for Mabor in Metical and Dollar values at the official rate. December 1986, 1987 are used as well as the trial balance for June 1988. 1l; addition the trial balance values are also translated into December !;o8 Dollars on the assumption that physical assets are basically unchanged. Table 6.43 HOZAMBIQUE - ASSET REVALUATION: MABOR DE MOZAMBIQUE Dec.1986 Dec.1987 June 1988 Dec.1988 Ht.H US$ Mt.H USS Mt.M US$ US$ Exchange rate (Mt/S) 40 400 450 620 Balance Sheet Assets Cash and Receivables 68.9 1.723 147.1 0.368 461.8 1.026 0.733 Stocks 304.4 7.610 572.5 1.431 1542.9 3.429 2.449 Fixed and Long Term 170.4 4.260 176.1 0.440 238.9 0.531 0.379 Total 543.7 13.593 895.7 2.239 2243.6 4.986 3.561 Liabilities and Net Worth Current 359.7 8.993 528.9 1.322 1174.8 2.611 1.865 Long Term 70.9 1.773 114.9 0.287 348.3 0.774 0.553 Net Worth 113.1 2.828 251.9 0.630 720.5 1.601 1.144 Total 543.7 13.593 895.7 2.239 2243.6 4.986 3.561 6.32. As can be seen fixed assets at book value in December 1986 were $4.3 million which can be considered as the real value of the depreciated assets. 14/ By December of 1987 these assets were worth only $440,000 and even at the overvalued trial balance figures by December 1988 they were carried at only $380,000 using the official rate of exchange. 6.33. "he revaluation of fixed asset law seeks to address this imbalance by correcting asset values by year of acquisition. The simple method used is for companies to take the present depreciated value of fixed assets and then 141 The trial balance figures are distorted by suspense accounts and other accounting mechanisms. For this reason asset and net worth values for June 1988 are overstated. - 100 - segregate them by year of purch&se or construction, applying a factor to compensate for the devaluations. The law provides for revaluation of completely depreciated assets only in cases where these assets can be shown to be in good condition and with a useful life. 6.34. The inflation coefficients, established by the law at the time of enactment were as followst Table 6.5: MOZABIQUE - COEIfICIENTS FOR REVALUATION OF PI7 D ASSETS Year Average Exchange Rate Coefficient (1988) until 1974 25.04 23.16 1975 27.24 21.29 1976 31.41 18.47 1977-80 32.55 17.82 1981 35.35 16.41 1982-83 38.98 14.88 1984-86 42.01 13.81 1987 (lst half) 200.0 2.90 1987 (2nd half) 400.0 1.45 1988 (lst half) 450.0 1.29 1988 (2nd half) 580.0 1.00 6. 35. For Mabor, the appropriate coefficient was 17.82, as the plant was installed mainly in the period 1977-79. As the June 1988 trial balance figures are overstated, the December 1987 asset figures (Mtl76.1 million) are a better base, which, excluding depreciation for the current year, would give the appropriate starting value. Reclassification of assets in the period 1986-87 causes some distortion; however, this is adjusted with an 80 factor, giving Mtl40.8 million. The company could also revalue completely depreciated assets which are still in good working order. 6.36. Preliminary estimate 15/ of fixed asset revaluation as of the end of 1988 was as follows: Estimated Net Asset Corrected Value* Factor Value value US$000 Mt.million Mt.million (at Mt620) 140.8 17.82 2,509 4.05 151 This estimate was guided by discussions with the companies themselves, who were not completely clear about how the law would be applied, and how fully depreciated assets should be measured in terms of remaining useful life and value to be used for correction. - 101 - 6.37. The above represents a simplifLed illustration of the impact of the law. The purpose of this calculation is Illustrative of the magnitude of impact on balance sheet values and is not intended to be totally accurate. the adjusted value would substantially alter Mabor's balance sheet. The estimated corrected value would represent an increase In assets and net worth 16/ of over $ 3.3 million. 6.38. The law as set out in 1989 provided that 602 of the new depreciation allowance can be used to offset profits taz in 1988; 75t in 1989 and 902 for 1990 and beyond. The law also provides that dollar denominated debt should also be corrected in a similar fashion. In the case of Mabor it still carries a $ 700,000 debt due to suppliers of equipment which will partly offset the gains in assets and net worth. 6.39. The law as Interpreted here is favorable to enterprises and allows for a substantial adjustment. The draft version of the revaluation of fixed asset law, over which It was reported that there was much discussion and some controversy, demonstrates the Government's continued commitment to the recommercialization and rationalization objectives introduced in January 1987. 6/ The off setting entry or "contra' account for the revaluation of assets will be a reserve account which can be considered as part of the net worth of the company. The law specifically states that this contra account should be used to adjust accumulated losses which firms may be carrying. - 102 - VII. STE PROMOTION OF MNDUSTIIAL INVKMTHMNT A. The Environment for Foreign Investment 7.01. In 1984 law 4184 (the law on direct foreign investment) was enacted. The law established the Office for Foreign Investment Promotion. 1/ Foreign investors are inter alia peLmitted to import capital and to repatriate profits, and fees. and the value of imported capital on sale or closure of an enterpriee. The range of incentives on offer are classified into three groupst those generally available to investors; those conferred according to specific benefits of th, enterprise; and those granted by Ministerial discretion for special cases. 2/ The law allows firms access to the domestic capital market on standard terms, although access to foreign exchange is limited. 7.02. Under the operation of the law capital invested prior to independence, in general, is not recognized as remittable. Profits derived from this investment are therefore not eligible for repatriation. ' This relates primarily to circumstances of the substantial capital fligc occurring prio.r to independence which was funded, to a large extent, by debt. However, the ownership rights of foreign capital are recognized and a wide variety of mostly Portuguese as well as some South African investments in Mozambique continue to receive direct or indirect supervision from overseas Head Offices. While these foreign companies have not made new investments, there has been considerable foreign currency expenditure on expatriate managers since independence. Many companies with substantial foreign ownership continue to employ expatriate managers. A variety of arrangements exist for financing the foreign exchange costs. In the cases of large conglomerates involved in import substitution production there are cases where the Banco de Mozambique has financed the costs. 7.03. In some cases Portuguese firms have absorbed part or all of foreign costs of managers without being able to refund these costs from the revenues of subsidiaries, because they considered the cost as an insurance, for the maintenance of a sizeable potential net worth of industrial op'rations in Mozambique. In the case of one investment of US$40 million, after initial difficulties with the Portuguese firm, the I/ The Gabinete de Promocao do Investimento Estrangeiro (GPIE). 2/ This information is set out in Investors Guide to Mozambique GPIE, August 1988, and Legislation on direct Investment in Mozambique, GPIE. 31 Certain exceptions to this rule were observed, e.g. in the case of Mabor where General Tire has received royalty payments since inception of production in 1979. However, these royalties are not equivalent to remittable profits on invested foreign capital. According to Ganeral Tire there have been no significant delays in receiving payments ($700,000 in 1988). Mabor of Portugal has also received substantial payments. - 103 - Government in 1986 negotiated a comprehensive management contract which recognized its rights of ownership and stipulated a Government option *buy out" in foreign exchange. Since 1987 the Portuguese firm has taken over the management of the industry and is optimistic about the prospects for profitability. 7.04. As shown in chapter 3 there are indications that foreign investors are increasing their interest in investments under the foreign investment law. In 1988 the largest group of investors, ir. terms of country of origin, was reported to be Britain, with some 352 of projects. Evidently British subsidiaries located in Zimbabwe have shown interest in the Beira corridor. Portuguese and South African concerns comprise 25? of proposals. * Portuguese investment has, for example, been proposed in cashew processing. 7.05. Indications given by existing and new investors are that the Government's policies on foreign investment are favorable but that there is a general attitude of caution, in particular due to the security problems. Once these problems are resolved the general opinion was that there would be substantial new investments by foreign concerns. B. Foreign Exchange Allocation: The Export Retention Scheme 7.06. The critical resource controlled by the Government is foreign exchange and, as discussed in chapter 5, this responsibility will probably continue for the foreseeable future. In industry the requirement is a) for timely, automatic and continuing access to foreign exchange generated; and b) access for those import substitution industries which are efficient users of foreign currency. 7.07. Interviews with the business community suggested that the export retention scheme is a major factor in the attraction of foreign investment. Its operation is therefore of key importance at least in the short to medium term (chapter 5 considered the longer run question of allocative bias). Since 1985 the scheme has allowed exporters more reliable access to foreign exchange for inputs. However, the retention accounts are denominated in local currency with no defined commitment to supply foreign exchange, and as such the value of the incentive to investors is obscured. 7.08. There are two main problems perceived by the investor; a) the possibility that, despite a valid entitlement, foreign exchange would not be made available when needed; and, b) ex"',ange devaluation risk for funds held within the _etention account. In the first case the difficulty may be one of overall foreign exchange scarcity which forces the Government to use funds from the retontion pool (Fundos Consigaados). This is less likely to be the case as the foreign exchange market approaches equilibrium. 7.09. The perceived devaluation risk relates to the time lapse between the date of receipt of a letter of credit for the foreign currency value of exported goods and date of opening of letter of credit for the value of imported materials. If a devaluation occurred in this period, in order to import planned quantitieis of materials, an enterprise would have - 104 - to find local currency bridging finance to purchase additional foreign currency at the new rate. The ultimate loss would be the interest cost on this financet however a loan would be liquidated through receipts at the next export cycle, and future imports would be automatically financed by *xport receipts. Thus the loss would be applicable only to the first export transaction, and would be of less significance as devaluation slows down. 7.10. Nevertheless interviews with businessmen suggested a percep'ton of a foreign exchange risk, which would act as a disincentive, particularly where large devaluati¢ns are expected. Thus, on the Government side a firm commitment to provision of foreign exchange funds, under the terms of the relevant company agreement, is desirable, and, provided that the company is financing its own foreign exchange through exports, and producing a net additionality, there would be a gain to the economy. In this context a revolving fund mechanism can be justified even for exports which have a high import content. 41 On pure foreign exchange grounds low local value- added activities could be justified provided that the earnings are clearly additional. However, monitoring and controls would be required. 7.11. One method of amending the scheme which could strengthen the incentive element is to give exporters automatic access to foreign exchange through the establishment of a BM special account in an overseas bank. A possible operating mechanism is given in technical appendix 4.1. The Government is not currently in favor of such an account; however, it has certain incentive advantages, as follows: (a) Instead of a book entry commitment held by the Government as at present, the commitment would be backed up by foreign exchange reserves. 51 (b) Participating companies could receive monthly statements form their foreign exchange subaccount, allowing them to plan the use of resources in a more systematic fashion. (c) The details of approvals and licensing of imnorts could be simplified. Participating companies with funds available would exchange local currency at the current exchange rate with relevant import documents and automatically receive the 4/ For example an exporting firm with 852 import content and with a neighboring market may be able to export 4 times per annual cycle. If each export earned $1 million the initial requirement of $850,000 in scarce foreign currency could generate $600,000 per annum in additional foreign currency. S/ This advantage is significant for foreign investors. For example, the Government made special concessions to LOHACO to conclude a foreign investment project, setting up an overseas account. The management stated that if a system as proposed here had been in operation the need for the special concession may not have arisen. It is possible that companies such as LOMACO would accept participation in the scheme proposed. - 105 - required foreign currency, which would be debited to the company subaccount. (d) Such a mechanism could provide a means for donor support of exporting companies. where an initial infusion of foreign currency 'seed capital" is required in order to undertake the first export cycle. The start-up investment is automatically replenished once the exporter is operational; and (e) The proportion of foreign exchango earnings which is retained by the Central Bank rather than by the exporting enterprises could be pooled as a central fund for import substituting companies requiring imported inputs, who might have access to a special subaccount. 7.12. A system of overseas sub-accounts for allocating hard currency could improve the efficiency of import substitution industries. A problem noted in many cases is the disruption caused in production by irregular import supplies under donor import programs, i.e. the tendency to go from "feast to famine to feast*. Many companies interviewed had a one or more year's supply of imported raw materials, resulting in high financial carrying costs. The system of sub-accounts could allow greater flexibility in purchasing of inputs than is presently the case. C. The Debt EQuity Swap Law: Effect on Investment Incentives 7.13. A large number of countries have enacted debt equity swap legislation in recent years, particularly in Latin America. The main objective of such procedures is (a) to assist in the reduction of the level of external debt, and (b) to add to the range of incentives available to investors with access to foreign exchange, thereby encouraging investment. There are a number of issues involved in the use of this type of instrument; these Include for example the possibly adverse consequences of major increases in foreign ownership of domestic assets, and secondly the additionality issue, i.e. whether swaps substitute for rather then add to new foreign exchange investments and therefore reduce foreign exchange inflows. Thirdly is the problem of the possibly inflationary impact of large-scale debt prepayments in local currency. 7.14. In the case of Mozambique however there is a special situation insofar as a significant stock of industrial assets already exists (over 100 intervened companies) which were previously non-Mozambican owned, and which are underutilized and have deteriorated because of the departure of owners who failed to transfer the necessary knowhow to Mozambicans. Provision of special incentives to acceptable foreign equity investors may carry the benefit of providing badly needed technical knowhow, with probable complementary financing, while at the same time returning assets to private hands, but according to criteria which accord with Government policy and preferences. 7.15. The Mozambique law, decree 24188, came into force in January 1989. The law allows for the legal transfer of external loan assets between - 106 - parties (e.g. creditor banks and investors). The foreign debt instrument would be cancelled on presentation to the Central Bank and replaced by an instrument denominated in local currency at face value according to the preva'ling exchange rate, for which a service fee would be charged by the Central Bank. Loral currency would be made available to redeem the Government's liability subject to an annual limit. The funds released would be investible in new or existing enterprises as part of their equity capital or for financing of local costs of project execution, or alternatively for investment in managed equity funds. Three conditions are of particular importance; a) article 5: - the eventual reversion of wnership to the state according to a given time frame; b) article 12: - restriction on final repatriation of the foreign exchange value of the holding (based on the original conwerted debt) to 20 years or more from the date of initial investment. c) article 8s - requirement for complementary additional resources in the form of machinery and equipment, technology, or foreign currency, to complement the funds realized through the swap, except in special circumstances under the foreign investment law. 7.16. Clarifications of the law in a number of areas may be required before it becomes a fully effective instrument; however, the three conditions stated above may be considered further from the point of view of their incentive effects. In each case they may involve a relatively stringent restriction on the potential return to an mvestor, by comparison with swap schemes set up in other countries. This is for the following reasons. (a) The reversion clause would apparently exclude resale of the investors equity holding in the market; instead the investor has to agree a valuation with the Government, and, without a clear title to the company, may not be prepared to develop long term plans for the enterprise. (b) The 20 year restriction on redemption of the original value of the investment raises a number of questions; as it stands this is a 'lock-in' clause which means that, whatever value of an equity holding is agreed with the Gover.ment, a significant part of the investors return would only be realized in the long term; a further problem may concern the rights to realize capital gains through retained earnings and increase in the real net worth of the enterprise over the holding period; finally, this restriction means that an investor cannot sell off according to his perception of business conditions; the 20 year rule is by international standards very restrictive; and (c) If the required complementary financing is valued at the official exchange rate, then the principal incentive to this type of investment would be diluted; this is because the investor is attracted to the debt equity swap route initially by the possibility of buying rights to invest in local currency at a discount through the secondary market for Mozambique's external debt. The larger the proportion of complementary funding at the official exchange rate the less favorable to the investor would be the effective exchange rate received on the - 107 - overall transaction..The overall effective rate is also depressed by Central Bank fees for the transaction. Finally, the effect of the parallel market for foreign exchange must be considered; if an investor can obtain all or part of his required Metical funds at a favorable rate on the parallel market, he will require an even more favorable effective rate through the swap route, especially if there are restrictions on the investment such as those under a) and b). 7.17. In contrast to the above restrictive clauses, under article 13 there is a guarantee of preservation of the *real values of invested capitai over the period of the investment. As it stands this would imply that redemption of the initial capital will be at the exchange rate prevailing on the date of original investment; this would mean that the foreign investor bears no exchange risk on the original amount invested, It is not clear how this would affect capital gains and dividends, but as it stands this is a geterous incentive, especially for enterprises which cannot protect themselves against currency depreciation through e.g. the export retention scheme. To some extent it would offset the disincentives under the 20 year rule. However clarification of the exact meaning of this article may be in order. 7.18. The immediate incentive to an investor of using the debt- equity swap route, as stated, is in terms of the favorable effective exchange which he can attain through buying discounted debt in the secondary market. A model of the effective rate calculation can be formulated. 6/ The example of note 6 assumes a discount of 80 on face value of the debt. However, the effective incentive to the investor is very much reduced in comparison what he would expect from the discount; an 80Z discount would yield prima facie a rate of five times official rate of exchange in Mt per US$. However this is reduced by the combined effect of the required complementary finance, and transaction costs to a rate about twice that of the official rate (an effective rate of Mtl,700 to US$1). 6/ Effective rate of exchange; Meticals per US$ 8 (F + C - T) X official exchange rate (D + C) Where: F = face value of dLbt purchased on secondary market (in US$); T - transaction cost; bank fees, legal/administrative costs etc; D - discounted price of debt on secondary market (US$); C = complementary finance valued at official exchange rate. Example: face value of Mozambique debt liability (F) = US$1.0 million value of debt on secondary market (20S, (D) - US$0.2 million transaction/fee costs to investor (10) (T) - US$0.1 million required complementary finance (50S) (C) - US$0.5 million Effective exchange rate would be 2.0 x official rate, which would yield approximately Mtl,700 = US$1.0 in January 1990. _ 108 - 7.19. The existence of a parallel exchange rate also has to be taken into account, if it is accessible to the investor, because it would further reduce the "effective" value of the discount incentive. For example, if the parallel rate is at Mtl,800 - US$1.0, then, on the above assumptions, an investor using the parallel market could get a more favorable exchange rate than if he used the debt-equity svap, while at the same time he would also avoid restrictive clauses such as reversion and the 20-year rule. In order to provide a sufficiently attractive incentive, on these assumptions, an investor with access to the parallel market would need to purchase debt on the secondary market at below 20? of face value. 7/ 7.20. The authorities need to carefully assess the financial impact of the various restrictions and costs placed on the scheme. If the ristrictions are so stringent that the investor requires excessively large discounts on the face value of Mozambican debt, then creditor institutions may simply decide not to supply debt assets and activity on the swap market would cease. On the other hand, given the alternative options available to Mozambique in terms of debt reduction, the Government may consider that it is justified to apply stringent restrictions, including the complementary finance requirement. However, the current law, by international standards, appears unusually restrictive and requires a number of clarifications. D. Industrial Promotion Infrastructure The Role of the lnidades de Direccao 7.21. The Unidades de Direccao or Management Units have played an important role in the past as monitoring and controlling organizations within the Ministries responsible for industry, 8/ overseeing both private and public sector enterprises. Under the most recent policy changes the Units are to be abolished and may be established as holding Companies, self-sufficient on the basis of dividend and fee income. A reorganization of the food and tobacco, and textile industry Units, is currently under 9/ implementation. A suggested approach to the reform of these institutions is as follows: 71 Mozambican commercial debt is trading currently at below 102 of face value. Despite this apparently favorable rate, it is still necessary to take into account the effects of complementary financing, transaction costs and the parallel exchange rate. The institutional infrastructure for industry is described in some detail by UNIDO; UNIDO (1987) op cit. These proposals are similar to those being considered for the new food industries company, previously known as UDRA, which is now being reorganized under the trade name XIGAIO, as a holding company, with a trading company affiliate. - 109 - 7.22. As holding companies these reconstituted units should oversee the specific investments which constitute their capital (i.e. Government investments), and exercise control in accordance with their holdings of voting shares of these companies. The formation of holding companies is appropriate as long as they are major investors in industrial enterprises. Holding companies would allow for a decentralized and specialized oversight of Government investments and should function as a means of consolidating certain services required by the constituent companies. In general, services would include: (a) Accounting and auditing: the holding companies should safeguard the interests of the investor (i.e. the Government) by ensuring that accounting and management information systems are functioning, ensure that the companies are well managed, and that the holding company itself has the proper tools to oversee its investments. Associated with this should be an auditing function to verify the accuracy of data. (b) Marketing: If the holding companies are to be subsector specific, which appears to be the intent, they would also have the ability to act as a marketing agency for a number of companies, domestically and especially for exports; and (c) Import services: import services were previously monopolized by Government entities who charged fees for the services. This monopoly has been considerably diluted with the advent of a large number of import-export agencies approved by the Government. However, the provision of import advisory services should still be a natural area of activity, with each Unit specializing in the specific needs of its company portfolio. However, there is a perception among constituent companies that, if holding companies control foreign exchange allocations, they will use their position to extract rents (e.g. excessive fees). This would involve a conflict of interest and should be avoided if the Units are to play an effective role in company support. 7.23. In order to avoid the proliferation of bureaucratic parastatal- type institutions a number of alternative approaches to the legal establishment o)f the new Unidades may be considered, the most important of which may be: (a) statutes of the companies could define the Units as autonomous entities managed by a board of directors, independent of Central Government. The board of directors would be constituted by managers or owners of companies with Government investments. Voting power of the holding company could be defined in relation to :he proportion of profits payable to the holding company as dividends. (b) the Units should be limited in their powers to those areas which are the interests of their constituent companies and do not derive from the powers of the Central Government. Powers of - 110 - price approval, foreign exchange allocation, and taxation should not be exercised by holding companies. (c) they should be self-sustaining entities based on dividend and fee income; and (d) they would only have the right to be involved in companies where there was some share of Government investment. The influence of holding companies would thus depend on extent of Government ownership, rather than sector of production as is currently the case. However, they may offer their services to companies which do not have any Government ownership or control on a fee basis. 7.24. The units are, presently, given an overall foreign exchange allocation based on requests by their client companies. The above proposals imply that responsibility for foreign exchange allocation would preferably be taken over by the Central Government. The Promotion of Small Scale Enterprises 7.25. The Government decision to change industrial policy in favor of small scale enterprise after the 1983 Congress (see chapter 2) resulted in the rapid creation of small industries units under seven Government departments. 10, IDIL (Instituto Nacional De Desenvolvimento Da Industria Local) was formed from the Department of Local Industry as an independent ncy under SEILA. In 1988 HIE and SEILA were merged, and IDIL became an ncy of MIE, charged with the major responsibility for small enterprise motion. In addition, in 1984 the Cabinete de Consultoria dos projectos Pequena Industria (GAPI) was founded as a technical assistance unit for __11 enterprise mainly in rural areas. ill 7.26. IDIL is responsible under its constitution for appropriate technology development and transfer, technical support and consultancy assistance, arranging of local and external financing, and mobilization of external cooperation, training, provision of advice on industrial protection; registration and data collection, and promotion of industrial activity based on local resources and needs. Most of IDIL's work has been in association with GAPI, and a few additional projects sponsored by other bilateral donors. On occasions IDIL has also despatched technicians to 10/ These included SEILA, MIE, Agriculture, Fisheries, Construction, Labor. The BPD set up a department specializing in small enterprise credit managed by GAPI. Ill This is supported by the F. Ebert Foundation of the German Federal Republic. GAPI has approved a total of about US$1.2 million to about 90 projects over 4 years (out of a total of 130 applications), and also arranges technical assistance, largely to small rural units, such as grain mills, vegetable oil, soap, woodworking and pottery mainly in Cabo Delgado, Niassa, and Maputo, and recently in Tete province. - ill - assist in field work, but its in-house technical staff, including its affiliate PRODIL which develops prototype village equipment, is minimal. Most recently, a small industry development fund has been established under IDIL to coordinate and channel aid funds. 7.27. The absorptive capacity of IDIL, as a main promotional agency for small enterprises, is severely limited at present because of its currently very small size and its reliance on Government budgetary resources. In addition, the number of external donors interested in assisting IDIL inevitably creates pressure on its internal administrative capacity. 7.28. It appears to be particularly important to examine alternative approaches to building up the agency as an effective instrument, but one which avoids the pitfalls of bureaucratic proliferation which have affected similar agencies in other countries. A key issue is the identification of the support needs of business, if any. 12/ The mere presence of a promotion office in a particular province does not necessarily produce any results, and could be counterproductive if the functions of the office are perceived by potential businessmen as largely ceremonial. Aside from a supportive macroeconomic framework, the assistance generally required by potential or current businesses involves a range of technical, managemnent, financial and infrastructural support. 7.29. Under the IDIL program a number of external donors have contributed in the area of training and provision of technical personnel and equipment. 13/ The World Bank is financing a pilot small business consultancy service, directed at providing management inf,rmation, finance, and marketing skills. This service could be used as an avenue for identifying the specific assistance needs of business. In this respect it should be noted that such needs vary considerably according to the existing state of kncwledge of an entrepreneur (e.g. ranging from preliminary ideas to an ongoing activity requiring specific inputs). A system of charging for services is also recommended, in order to permit a degree of independence from public funds, and allow autonomous management decisions. 7.30. A possible option for the development of technical capacity in IDIL is to set up a *skills exchanges, to subcontract technicians from existing larger industrial enterprises for provision of assistance to smaller enterprises. This proposal would have the advantage of integrating existing skills, where available, into the overall recovery effort and at the same time reducing the budgetary and administrative implications of a build-up of IDIL inhouse staff while allowing expansion of technical 12/ A number of similar agencies elsewnere in Africa appear to have been developed from standard models (e.g. from India), rather than as a result of appraisal of the specific needs of small businessmen in the particular country. Lack of a demand-driven approach has inevitably encouraged bureaucratization. 13/ These are: UNIDO, ILL, SIDA, USAID, VITA (US), DOG (Holland), HIVOS (Holland), and the Eduardo Mondlane Foundation. - 112 - capacity. Whatever options are adopted, it is in any case important to direct attention to the issues of: (a) identification of real needs of small businessmen; (b) cost effective forms of assistance; (c) cost recovery for the assistance provided. S. Development and Diversification of Industrial Financing The Environment for Equity Investment 7.31. The financial sector is likely to play an increasingly important role as rehabilitation proceeds. As shown in chapter 6, distortion of the capital market has led to a severe reduction in the role of the banks, including concentration on short term overdraft financing. The banks have not participated in fixed investment financing for some years. due to a combination of interest rate repression, lack of foreign exchange, and disruption of investment plans in an increasingly risky environment. The design of the future financial sector consequently presents numerous possible options. Financial deepening would be associated with the establishment of diverse types of financial instruments, including a range of deb*t, equity and quasi-equity instruments, lease-financing etc. This section focuses on the possible options for generating equity funds. 7.32. New and rehabilitated enterprises will need substantial long- term finance, in terms of local and foreign currency. However, the major replacement requirements of some companies and low current profitability mean that funding rehabilitation through debt finance involves burdening companies with high fixed charges. Which could create cash flow problems in early years of rehabilitation resulting in insolvency and renewed debt service problems for the banking system. There would probably be significant benefits if a way could be found to finance investments through equity, whicti reduces capital payments early on in a project life, as a complement to the intended revitalization of bank lending. 7.33. The Government is interested in the sale of enterprises presently under its direct ownership or intervention, if possible on a joint venture basis, or some other basis where the state can maintain a stake in the enterprise. 14/ Certain intervened industries have been also reclassified as private and in some cases the Government, which had control o ir a majority of shares from departed owners, sold its participation to the minority remaining owners. Under the 1989 law (see chapter 3) the Government has clarified the legal position of intervened companies in order to allow full or partial divestment. This environment of ownership transfer is one in which the development of equity financing may be regarded as important. 141 Recently, one state company in the construction sector (PROSUL) has for example proposed the sale or rental of approximate'y half of the production units which it oversees. - 113 - Developmmnt in Financial Services and Markets 7.34. Two initiatives taken in 1988 included the formation of a capital participation department in BPD to manage its equity portfolio, largely in the textile industry, and the creation of the Sociedade de Investimentos e Estudos Financeiros (SOCIEF), which was established to provide a secondary market for Government bond issues, but has acquired a wider terms of reference. The principal shareholders of SOCIEF are state corpoiations, however private sector corporations are also represented. 15! Some shareholders contributed capital in the form of Government bonds, so that SOCIEF is also involved in the market on its own account. SOCIEF covers two broad areas of activitiess (a) underwriting the issue of shares and other securities allowed by law, and acting as the market agent for government securities, and; (b) provision of consulting services for companies, particularly feasibility studies, management consulting, restructuring, reorganization and privatization 16/ of companies. 7.35. SOCIEF may be regarded as the precursor of a securities marketing service, as it is a mechanism which can fatAlitate the transfer of asset ownership in a timely and efficient manner. While not expected to act as a holding company, SOCIEF is empowered to own shares in other companies and may be in a good position to trade in securities once it creates the technical expertise required for the services it is already comnitted to offer. The BPD capital participation facility is one alternative vehicle for the development of equity investment. 7.36. The limited availability of risk capital in the private sector to establish an investment and financial services company, and the extent of existing Government ownership of industrial assets via the state and intervened companies, justifies state sector control of an investment facility, at least in the short term. However, it would probably be advisable to establish it as a fully independent agency, linked to institutions such as BPD, capable of being 'spun off' later, possibly with private sector involvement. The development of securities marketing capability should provide the mechanism, inter alia, for; (a) raising capital (through corporate stock/share issues etc), (b) sales of stock and bonds by investors wishing to exit from the market or restructure their asset holdings, and (c) providing a ready outlet for savers wishing to place savings outside the banking system, either directly in new productive assets, by purchase of existing stock, or through managed equity funds (e.g. unit trusts). A reasonably active market is necessary to generate 15/ Three state corporations own over 80Z of shares: Banco Popular de Desenvolvimento (44?); Empresa Mocambicana de Seguros (23Z); and Linhas Aereas de Mozambique (162). The private shareholders include Banco Standard Totta and Cia. Joao Ferreira dos Santos. 16/ The term privatization (privatizacao) is not used in the company statutes. However the term reconversao (reconversion) is used, which implies returning companies to private -actor ownership. - 114 - interest in this outlet for savings. In addition, securities marketing centers play an active role in the buying and selling of Government bonds. 7.37. One possible constraint on the creation of an effective securities market would be a Government inclination to maintain long term rights of ownership; this is the implication of the restrictive clauses (e.g. the reversion clause) in the recent debt equity swap law (discussed above). This Issue may require further consideration by the Government. Resources of Entrepreneurship in Nozambique 7.38. Equity investment assumes the existence oft i) entrepreneurs who, if assisted with investment capital, would have the ability to manage startup or existing enterprises successfully; and ii) commercial opportunities either in startup or the takeover of existing enterprises. m From chapter 4 it is clear there are viable investment prospects within the industrial sector at least in terms of rehabilitation of existing companies. More difficult to define is the future source of entrepreneurs. The legacy of the colonial era, along with Government policies in the late 1970's and early 1980's, has influenced the evolution of entrepreneurship. While the private sector during this period was tolerated it was not given 'ive encouragement and Hozambicans receiving university training and erested in business were essentially limited in choice of employment, tly joining the public sector. Leaving aside foreign investment, small ders and the informal sector there may be considered to be three basic -J.es of entrepreneurs X Mozambique: (a) Industrial Sector Entrepreneurs. Consisting largely of owners who remained in country after independence, this class of entrepreneur has survived but has not accumulated capital in a significant fashion due to pricing policies and the other problems up to January 1987. These entrepreneurs represent a significant existing resource. Of 575 nationally-registered industries, close to 300 are private sector companies. (See chapter 2). The owners of these industries are facing, in general, low capacity utilization and the full range of rehabilita.;on needs. Equity investment would be welcomed if it could be accomplished without influencing the management or control of these companies. If equity investment were to be undertaken by a state or parastatal financial corporation it would be important that the shares be non-voting, or some form of preferred shares. (b) Commercial Sector Entrepreneurs. The small private trading sector is widely reported to have benefited from the price distortions prior to the ERP. This has allowed a substantial accumulation of capital. However, according to international experience, the transit'An from trading to managing productive industry is not necessartly straightforward, so it is difficult to assesc the potential in this type of entrepreneur. Government officials also appear to be rereluctant to consider these entrepreneurs for the sale of existing industries as - 115 - there have been cases where irdustries purchased were stripped and the facilities used for storage space or additional retail outlets. Given the past history of accumulation these entrepreneurs may not be the most likely equity investors; and (c) Public Sector Entrepreneurs. While the concept "public sector entrepreneurs" may seem a contradiction in terms it is justifiable in the context of Mozambique. The considerable manpower which has been employed in the public sector to manage various enterprises or management units since independence constitutes a pool of potential talent. Clearly not all of this talent could be considered as entrepreneurial but a small percentage could yield a significant number of potential entrepreneurs. The on-the-job practical experience of the past decade may have been significant in forming good managers. The size of the pool can be estimated by looking at three groups of managers in the public sector: management units, state companies and intervened companies. 7.39. To maximize the use of limited manpower in 1979 the Government grouped private, intervened and state companies under the Unidades de Direccao by subsectors of operation. According to 1986 data there were over 300 staff employed in 11 of these Units. 171 Of this staff close to 100 were mid to senior level managers. 7.40. As stated in chapter 2 there are some 114 state companies which vary in size from small operations to conglomerates employing thousands of workers. With an average of four mid to senior level managers per company a pool of about 450 managers may exist. A similar number may be available in the intervened company sector, which in 1987 consisted of 140 companies at national level. (See chapter 2). In addition some enterprise managers were transferred to Government service. Out of approximately 1,000 individuals in the state sector with management experience potential entrepreneurs may exist who would form part of the market for equity funds. The Industrial Labor Force: Share Participation Schemes 7.41. In addition to entrepreneurs another possible source of shareholding is the industrial labor force, through profit sharing or equity participation schemes. Equity ownership of this type could create incentives to productivity; however schemes could also permit some form of participation in management decisions-or policy. The attraction of such schemes is that they *.ould build on the historically relatively high level 17/ This information is provided in UNIDO; op cit (19P7). The 11 Units existing at the time were only in those subsectors which were under the responsibility of the Secretariat of Light Industry and MIE, totalling 419 industries. Additional Management Units existed for the oversight of some 156 industries under the responsibility of other Ministries. - 116 - of involvement of the labor force in Mozambique, which originates in the attempts to reactivate production after independence. 7.42. Over 1976-80 the work force became politicized through the 'grupos dinamizadores". Interviews suggested that during this phase managers were intimidated by labor and the efficiency of production and management suffered. Production was often disrupted by political meetings; at the same time this period was also one of partial recovery in the industrial output (see chapter 3). Over 1981-82 the creation of the Organizacao dos Trabalhadores Mocambicanos as the overall umbrella labor organization and a depolitization of the work force resulted in managers assuming more control. Since 1983 the work force has continued to evolve so that at present, in many cases, the "grupos dinamizadores' appear to be functioning like 'quality control groups". The worke- groups have thus progressed from political mobilization to production motivation whereby workers discuss how production quotas can be met, problems in pru.uction and suggestions for improvement. 7.43. Expatriate managers with experience in various countries in Africa observed that the Mozambican labor force was relatively hard working and well disciplined, although the availability of training remains poor. Local managers considered that the present level of cooperation with labor was relatively good. The quality of the labor force is a major determinant of the comparative advantage of Mozambique and such a resource is of particular importance in the attraction of foreign investment and undertaking of competitive production vis a vis third World countries of, for example, Asia. 7.44. On the basis of experience in a number of other countries, there is a case for introducing a system of "Employee Stock Ownership Plans" (ESOPs). An equity investment scheme may increase motivation through facilitating the purchase of shares by workers, enabling them to share in the success of a company more than is presently possible through salaries or bonuses. In return for an expectation of increased earnings as a result of higher labor productivity, principal stockholders of an enterprise may consider that there would be justification in transferring a proportion of equity to the labor force. To operationalize such schemes a series of models need to be worked out in detail, in terms of worker eligibility, the proportion of equity transferred, its price and possible options for payment (if any), phasing, liability, board membership and voting, exit (resale), dividend rights etc. It would be advisable to experiment with pilot schemes initially in order to determine how this instrument could be used to increase productivity, which would be its main objective. The success of such schemes however is dependent on a favorable environment. Thus it is important that the types of reforms proposed in this report are in place, or under implementation. The Role of Venture Capital 7.45. In its classic form venture capital is used by owners of capital to complement entrepreneurs who own patents and rights to innovations. Thus venture capital is normally associated with risky, high technology products at an early stage of their life cycle. Equity - 117 _ investment or venture capital companies in developed countries are normally started up within a capital market where shares can be sold once the venture proves successful. This allows investors to spread their risks among a number of veatures. In most recent times the main areas of venture investment have been in computers, electronics and bio-technology, some of which have shown explosive growth and profitability. 7.46. In Britain it is estimated that in 1988 about US$7 billion of assets were managed by about 100 firms which could be classified as venture capital companies, ranging from the largest with about 100 staff to those which employ one or two professionals. In the USA 600 firms manage about US$16 billion worth of assets. Experience in the developed world has shown a 602 failure rate from new ventures of this type, and a 202 rate of highly successful ventures. Typically therefore an equity fund of $10 million which invested $1 million in each of ten industries would have six failures, two which maintained the! -alue, and the remaining two which would need to appreciate sufficiently in value to generate a return high enough to make the overall return on the original investment worthwhile. This would mean that the value of investment in the two successful companies would have to appreciate on the securities market from $2 million to $10-12 million. 7.47. This model is however not applicable in the context of Mozambique or most developing countries. There is at present neither the stock market nor the capital available from many investors in Mozambique or other developing countries to allow the type of appreciation normally associated with venture capital. Where the securities market is undeveloped equity investors can generally only expect two basic sources of return or valuation: (a) dividend income and (b) entrepreneur buy-out of the investors shares. Another fundamental difference in Mozambique would be the absence of innovation as the basis for partnership; instead it will be the ability to manage which will probably be critical. This is not to say that Mozambican entrepreneurs with good ideas and innovations may not appear. but at best this will be a small part of the potential market for equity investment. 7.48. The main use for equity investment in Mozambique would be as an instrument for financial restructuring or privatization of state-owned or intervened industries, taking into account the need to safeguard the Governament's share of the sales proceeds of assets which appreciate over time. In this respect it is probably desirable that state equity investment should be on a "silent partner' or 'preferred share' basis, so that only if companies do not fulfil agreed performance conditions would the state have the right to intervene in management. The promotion of equity investment would have the following objectives. (a) As a means for transferring the management, control and ultimately ownership of public sector indtustries to competent management. (b) As a possible means of providing greater incentive to workers by allowing ownership and participation in profits. - 118 - (c) To strengthen existing private sector companies without influencing their management or control as long as these firms perform within agreed guidelines; and (d) To provide a return to investors which should equal or exceed, over time, the interest paid on long term debt. Nanagement, Legal, and Accounting Services 7.49. Since 1975 the legal profession has been effectively nationalized; at the same time accounting was tied to the needs of the central plan, and external company auditing essentially disappeared. Without the necessary skills in these areas investment capital operations would be difficult. Equity capital financing requires accurate, timely, meaningful and independently verified financial information on past and ongoing company performance. It also requires a legal system and services to safeguard the interests of involved parties. 7.50. The present environment for these services has improved. Starting in December of 1984, accounting took on a more important role with the introduction of a comprehensive accounting system for all companies. While the objective of this system was for better centralized planning and orientation of the economy, the balance sheet plan of accounts is fairly standard and allows for conventional analysis. However, the profit and loss accounts appear less practically applicable. 18/ 7.51. The system of accounts is designed more for purposes of tax assessment than for use as a management information eystem. However, local professionals state that the data presently available in companies can be used to construct proper management reports. Effectively companies are 'literate' but have +o become Oanalytical'. This, it should be noted, is far easier in many respects than attempting to introduce a completely new basic accounting system. Managers of many enterprises (public and private) are well versed in unit cost analysis for the purpose justifying price increases. Reconciling the estimates used in these analyses with operating data would be the task of a management information system. 7.52. While there are no auditing firms operating in Mozambique there are two private multi-service consulting firms which offer accounting services and could be developed into auditing firms, possibly as representatives to internationally recognized firms. With the increased influx of donor aid several international firms from Portugal as well as the *big eight" have been working in Hozambique, in non-auditing areas, and the prospects for establishing auditing firms seems to depend more on the demand for this service rather than availability. However, at present, independent auditing is not required in any systematic fashion outside tax assessment. 18/ This observation has been made by Arthur D. Little and Coopers and Lybrand in discussing the lack of management information available in a variety on industries interviewed. - 119 - 7.53. Inevitably, given the historical situation, management skills in the industrial sector are weak. 19, Technical assistance is required especially in the areas of management information systems and marketing. L'cal consulting firms are beginning to emerge to fulfill these needs and the prognosis for improvement in management levels is good given their increased accountability in terms of results. For the appraisal of equity investments specialized services to undertake business analysis, valuation of assets, profitability projections etc are also required both for project approvallrejection and for ongoing monitoring. With the improved environment for accounting, auditing and legal services new firms are appearing potentially capable of offering these services to entrepreneurs. 7.54. Finally, with respect to the legal environment, Government commitment to the rights of private ownership appear well established. While lawyers are still not an independent profession it also appears that some moves in this area are likely in the near future. Currently a review of the corporate legal framework is under way by a Governmental working group. 20/ F. Hechanlsms for Privatization and Transfer of Companies 7.55. Selling State-owned or intervened companies when the potential buyers do not have the necessary capital is complex and would require special legal instruments which safeguard the interests of the various parties involved. When, in addition, these is also the need for recapitalization and financing for rehabilitation the instruments and agreements become even more complex. The agreements and instruments would need to guarantee in particular the followings (a) That the State, over tim, receives a fair price for the assets. This price should be determined by the following main considerationst - The profit-earning potential of the assets; - The value of outstanding liabilities; - The additional equity investment required. The mix of fair market value and potential profitability considerations can only be assessed on a case-by-case basis, with two distinct valuation approaches used, based on (a) some estimate of the selloff value of the fixed assets and inventories, with allowance for accumulated cash or debt 191 This problem is highlighted in the Arthur D Little survey of July 1988. In several parts of the report covering 40 large industries mention is made of the inadequate management capacity. 201 The World Bank is also carrying out a study of the corporate legal framework focusing on ownership, and sale of enterprises, incentives, auditing, employment legislation, and banking regulations. - 120 _ fixed assets and inventories, with allowance for accumulated cash or debt (i.e. the revalued net worth of the company balance sheet at the date of purchase), and (b) an estimate of the value of future cash flows of the existing company. Large state corporations which are currently unprofitable will, for the most part, not be easily saleable. The additional equity requirement would be an additional element in the price, and there will also be cases where the needed new capital will be the predominant consideration. (b) That the potential buyers have the freedom to properly manage the companies and that their rights to purchase are clearly stated. To accomplish what has been set out two basic contractual arrangements or instruments may be considered. These are as follows: (a) Transfer of Company Control; Preferred (non-voting) Shares 7.56. The concept of preferred or non-voting shares existed prior to independence. For decentralization and privatization of public sector industries it could be used as an instrument for transfer. Preferred shares, as proposed here, would be a flexible instrument which would safeguard Government interests (the value of its shareholding) but would not involve Government control of the management of the industries. 7.57. In order to achieve this, existing Government ownership 211 could be converted into shares which have no voting power and, for a specified period, no rights to dividends. New non-Government equity investment, which would have voting rights, would also have controlling rights. 221 In the case of sale of the company by the voting shareholders the preferred shareholders would however retain the rights to their proportionate share of the proceeds. In the case of a 1002 buy out' (e.g. privatization) where ownership is changing hands, existing rights to dividends of Government's preferred shares would be car.nelled or curtailed. The relative rights of the two types of shares would be negotiated on a case-by-case basis. 7.58. This concept of shareholding is unorthodox but is a possible instrument for transferring ownership in Mozambique's particular 21/ This should include intervened companies which are not technically owned by the Government. According to the 1989 law, an intervened company has to be nationalized prior to sale, and then the shares are sold to the intended buyers. 22/ In this respect it would be possible for new voting shareholders, even if they are in a minority. to effectively control the company as long as this is clearly defined under the restructuring agreement. - 121 - circumstances. 231 The use of preferred shares as proposed would ensure that the Government retained a long-term stake in the profits of the industry. The proceeds of any sale of the industry would be divided in direct proportion to ownership of total shares. Thus, if the holder of preferred shares constituted 60T of invested capital, they would not have the rights to 60? of profits (dividends) but they would have the.rights to 60S of the sale proceeds if an enterprise was liquidated. Finally, in cases where rights to dividends are allowed on preferred shares. they should normally be determined by the market value of the company's assets and profitability prior to the new investment. However, where the profitability of the enterprise after transfer is particularly high, it may be necessarf to allow some rights to dividends to preferred shareholders ex post. (b) Transfer of Company Control via Management Contracts 7.59. A further instrument for facilitating transfer of ownership and control could be a type of management contract. This would be designed in such a way as to clearly define the rights and responsibilities of the Government on the one hand and the entrepreneur(s) on the other. In legal terms the entrepreneur could be an incorporated company or a partnership. 24/ A partnership arrangement might be more flexible in terms of thare acquisition, taxation, reinvestment of profits. hiring of technical services and loan applications, because decisions do not have to be referred to a board. Included within the management contract would be two main areas of agreement: - rights of management, and terms and conditions for purchase of a company(s). Technical appendix 6.? contains a preliminary outline of what the basic elements of such an agreement could look like in a hypothetical case involving a partnership, with other features including a worker participation scheme. G. Institutional and Management Considerations for Equity Finaneing 7.60. Equity financing for developing countries has become, in recent years, a growing area of interest and funding by donors. The principal problems which have affected other schemes are the following: (a) Legal Environment. Many developing countries do not have a legal framework which allows for the type of flexibility required for equity investment. 231 Normally, preferred shares have no rights to control but would have equal and priority rights for dividends, as well as participation in sale proceeds. Tre waiving of these rights would be an extra incentive to irvestors required in order to compensate for the uncertainties involved, (it would not therefore imply a subsidy). 24/ For the purpose of discussion only the term 'partnership, will be used. In the case of public sector entrepreneurs it would be advisable that up to four managers formed such a partnership to ensure the mix of skills required for managing companies. - 122 - (b) Valuation of Shares. Equity investors compensate for inevitable losses through share value appreciation from a few successful companies. The legal mechanism, or stock market, in which shares find a value is often inadequate. This is a problem of the "exit scenario" for equity owners. (c) Income prospects. Because of the valuation problems the income or appreciation of investments will be derived solely from dividends and buy-outs by owners/managers of companies. Estimating valuation or future price for purchasing shares is a complex task which would have to be accomplished prior to investment because subsequently conflict of interest could develop between owners and outside shareholders. (d) Reliability of reporting. The entrepreneure may have rn incentive to 'legally" hide or disguise the profitability of a fimr, thus minimizing dividend payments (and valuation prospects). This is especially the case with export/import companies. External auditing has to be comprehensive, ongoing and along strictly enforced guidelines. The auditing practices and traditions in developing countries are often inadequate; and (e) Cost of fund operation. In the case of an equity fund the costs of management can be high, particularly if expatriate expertise is required. At the same time equity may be perceived as a grant by entrepreneurs, resulting in a transfer of wealth (assets). This problem has occurred for example with schemes which benefit former civil servants. Experience to date has shown that technical services (consultancies) required to formulate an indepth proposal or prospectus will have to be subsidized in whole or in part. 7.61. Two additional and fundamental areas of concern require in- depth analysis: a) the possibility of establishing adequate auditing capabilities and the costs and, b) financial viability of an equity financing scheme. Assuming these issues can be effectively addressed, along with the other problems, then a hypothetical equity financing scheme could be possible. 7.62. Previous experience seems to indicate that banks are not the most appropriate institutional mechanism for equity financing as the approach to analysis and ongoing monitoring is substantially different to lending. Banks tend to be collateral-oriented while equity investment requires a risk-orientation based on the future prospects of the intended investments. Nevertheless, as stated, BPD has taken formal steps to move into this area. 7.63. Alternative options for an equity financing institution involve either creating a new institution or expanding the scope of existing ones. - 123 - A new entity would probably be established under a board of directors. 25/ Staffing of the new entity would be limited as indepth appraisal work should be conducted by the entrepreneurs with assistance, as necessary, from professional services available in country or from external sources. The core staff should have the technical ability to review and verify proposals, but primarily their function would be to restructure, liquadate and transfer companies. The board should be constituted to supervise overall management and approve investments according to operating procedures. 7.64. The basic functions of the new institution would be twofold: (i) as a direct equity investor; (ii) as a *packager, which assists in leveraging its investment with additional risk capital (local and/or foreign) as well as conventional debt finance. Function (ii) could include the design of enterprise-specific financing schemes, including recapitalizations, sales, acquisitions and mergers. The institution could act as a 'magnet* to a variety of sources of debt and risk capital and act as an intermediary to assist in negotiations with the banks, the foreign investment board and additional investors. The professional and management skills required for a 'fundw would include: (a) investment fund management; (b) accounting/auditing; (c) corporate finance; and (d) legal expertise. Additional technical expertise for engineering studies, marketing, should be contracted as needed. 7.65. The import-dependent nature of the economy vould imply that an investment fund bP capitalized by foreign currency resources. These resources, at present, would be available largely from the donor community. The industrial credit survey suggests an outstanding demand for funds well in excess of US$100 million, of which debt finance would comprise only a part. There is a good case for donor finance via grant funds for the initial establishment of an equity investment facility, if such a facility provides an efficient channel. In addition, preliminary discussions indicate that international companies and investment banking institutions might be prepared to provide initial funding. Sources of management expertise for such an institution could also be provided from the same sources, as well as from specialized investment institutions. If the proper professional skills are recruited and put in place, and necessary resources provided, the equity fund could act as an important catalyst, providing: (a) an important means of attracting/leveraging new foreign investment; (b) means of recapitalizing existing private sector industrial sector firms which will improve their viability by reducing debt burden; and (c) the means by which the Government could divest itself from industries it currently owns or controls. 25/ Other models are however possible including limited partnerships which are more comon in the USA; the establishment of a board is more common in Europe. - 124 - B. Conclusions 7.66. The present environment is one in which forms of industrial finance which do not burden infant or newly rehabilitated companies, should be used, alongside conventional bank lending. Enterprises need access to non-debt financing. In order to divest intervened and, possibly, state enterprises, in a way which maximizes the mutual benefits of both Government and potential purchasers, the Government should promote an institution spocializing in equity investment capable of dealing in company transfer and restructuring. In order for the institution to be effective simultaneous development of instruments for transferring ownership and control need to be developed, along with, over time, securities marketing facilities and the infrastructure needed to support securities markets (in particular a regulatory system, accounting and information infrastructure). 7.67. Close monitoring of the impact of the fiscal and monetary incentive system on local and foreign investors is also required, and active steps to promote indigenous entrepreneurship. The institutional requirements are being put in place, albeit in preliminary form, and existing investment institutions should be diversified in order to provide some of the necessary financial services infrastructure. _ 125 - Technical A#nendiz 4.01 Pagel of 9 Domestic Resource Cost (DRC) Measure of Economic Efficiency 1. The DRC ratio is used as an indicator of whether a country has a comparative advantage in particular lines of production. It compares the domestic resources used in production with the net foreign exchange effect of that production, aud can thus assess the relative merits of earning foreign exchange through additional exports, or saving it through import substitution. DRC ratios can be used to assess the absolute efficiency of activities - the criteria here being whether the cost of additional foreign exci.ange in terms of domestic resources is less than the value of that foreign exchange to the economy. However, in practice, uncertainties regarding the value of key parameters, including for many countries a precise estimate of the shadow exchange rate (SER), mean that absolute measures of efficiency may be less helpful than the ranking of activities by their DRCs. l/ Definition of the DRC ratio 2. Formally the DRC ratio compares the full domestic costs of production (domestic labour + domestic content of capital and non-traded inputs) with the net foreign exchange effect (output at world prices - traded inputs at world prices - foreign exchange content of capital and non-traded inputs at world prices). The approach adopted here is the simplified one of defining domestic resource costs as direct labour inputs plus all non-traded inputs, (including, if data are available, any capital costs which can be identified as non-traded). There is no attempt to decompose non-traded inputs nor to shadow price labour. The denominator - the net foreign exchange effect - is the value of output at world prices minus all traded input costs at world prices, where all values at world prices are converted into metical at the official exchange rate. 3. Formally, in terms of a unit of output of good i, DRCi can be expressed as: DRCi * a, + 21.4DP. 2/ (Wpi - X aji WPj)I OR I/ Theoretically it is not strictly correct to rank on the basis of DRCs, however, in practice any errors arising from such a ranking are not likely to be great. 2/ Any foreign exchange costs of labour - for example remitted salaries of foreign workers - will be included with a negative sign in the denominator. - 126 - Technical Appendix 4.01 Page 2 of 9 where ali is the number of workers employed in the production of i; VP is wage per worker; ani is the number of units of non-traded input n required in the production of i; DPn is the domestic price of n; aji is the number of units of traded input j required in the production of i; WPj is the world price of a unit of j; OER is the official exchange (MT per US$) and summation signs indicate totals for all labour, non-traded and traded inputs. 4. This form of the DRC is given as a whole number, with the normal efficiency condition being a DRC of 1.0 or less. However, where the official exchange rate does not reflect the value of additional foreign exchange to the economy, so that there is a shadow exchange rate that differs from the official rate, the efficiency condition will be that DRC is less than SERIOER. For example, where the SER is 50Z above the official rate, efficiency requires a DRC of less than 1.5, not 1.0. In extreme situations the DRC ratio can be negative, where the foreign exchange value of output is less than that of the traded inputs used to produce it. Such activities generate net foreign exchange losses, in addition to using domestic resources so that there can be no value of foreign exchange at which they will be economically efficient. 3/ 5. In practice, where there is uncertainty regarding the real value of foreign exchange it may not be possible to do more than specify a range within which efficient activities are likely to be found. For example, using a range of 1.0 to 1.5, efficient activities will have DRCs below 1.0, and unambiguously inefficient ones DRCs above 1.5. Those with DRCs within the range may be efficient depending upon how one values foreign exchange 41. The specification of the DRC measure used here differs slightly from that which has been estimated by some Ministries in Mozambique. There the tenmr in equation (1) are modified so that only labour is included in the numerator as a domestic resource cost. All 31 However nontraded output in industries such as bulky building materials would have to be treated with a modified version of this calculation in which a foreign exchange equivalent value is used for output. 41 In this study for illustration it has been assumed that foreign exchange is overvalued and that the 'real' rate of exchange should be 1.30 x the official rate. - 127 - Technical Appendli 4.01 Page 3 of 9 non-traded inputs are treated as equivalent to foreign exchange costs and subtracted from the value of output at world prices in the denominator. Equation (1) thus becomes: DRCI - £ ali Vp _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _(2 ) (WPi - aji Pj) * OER - E ai DP, 6. Both equations (1) and (2) are simplifications in that strictly what is required is to decompose all non-traded inputs into their labour and traded good costs. Equation (1) assumes implicitly that such non-traded inputs are composed totally of domestJ.c labour, whilst equation (2) makes the other extreme assumption that they have a zero labour content and are all traded costs. Which is the more accurate assumption will vary with the cost breakdown of the non-traded inputs concerned. It is felt that the assumption underlying (1) is more reasonable, for non-traded activities in Mozambique. 7. Normally whilst use of equations (1) or (2) will give differenf DRC figures, provided the same value is placed on fore-gn exchange in both calculations (either OER or an estimated SER) an activity which is efficient by one equation will also be efficient by the other. A serious difference between the two equations will only arise when there is a weak foreign exchange effect - so that the difference between output at world prices and traded inputs - is positive, but low. In these circumstances use of equation (2), which includes non-traded costs in the denominator, can give a negative DRC, which can be misleading, since a negative result normally Implies a net loss of foreign exchange. 8. Use of equation (1) in these circumstances will give a high, but positive DRC, so that the activity will appear inefficient, but not a net loser of foreign exchange. In theory, whenever an activity has a high, but positive DRC ratio, there will be some value of foreign exchange (or ratio SERIOER) at which it becomes competitive. Where there is a negative DRC, however, an activity can never be competitive. It is desirable to ensure that only activities with genuinely negative foreign exchange effects have negative DRCs, so that it is suggested that equation (1) is a preferable formulation than (2), although both are approximations. 9. Table 1 illustrates the use of these alternative equations with a simple example. There is a single output A. and two inputs B (traded) ani C (non-traded). In case 1 the activity has a strong positive foreign exchange effect, since at world prices converted at the OER output is 100 and traded input B is 50, giving a net foreign exchange effect of 50. In case 2 this foreign exchange effect whilst still positive, is weaker, since output is now valued at 60, giving a net effect of 10. 10. In case 1 both equations give DRCs of less than 1.0 implying efficiency at the OER. However, in case 2, whilst equation (1) gives a high DRC of 4.5, implying significant inefficiency, equation (2) gives a - 128 - Technical Aipendiz 4.01 Page 4 of 9 negative result, because of the negative denominators This is despite the fact that the actual foreign exchange position is not negative. Table 1i KOZANBIQUE - ILLUSTRATION OF ALTER&TIVE DRC EQUATIONS Case 1 Strong Foreign Exchange effect I~~~~~~~~~~~~~~ World Prices 8 Domestic Prices Output A (Traded) 100 120 Input B (Traded) 50 75 C (Non-Traded) 20 Labour 25 Equation (1) Equation (2) DRC - (25 + 20) - 45 DRC - 25 -25 (100 - 50) 50 (100 - 50) - 20 30 3 0.90 m 0.83 Case 2 Weak Foreign Exchange effect World Prices a Domestic Prices Output A (Traded) 60 120 Input B (Traded) 50 75 C (Non-Traded) 20 Labour 25 Equation (1) Equation (2) DRC - (25 + 20) 45 DRC 25 -25 (60 - 50) 10 (60 - 50) - 20 -10 - 4.5 e -2.5 Notes a) Foreign exchange value converted to the ORR Short-run and Long-run DRCs 11. A comon finding of DRC studies is that the results are sensitive to the current level of plant operation and to the treatment of capital costs. It is necessary therefore to specify whether what is being estimated is the efficiency of current operations - given by a short-run DRC - or of production in the longer term - for which a long-run DRC is required. This distinction between short and long-run DRCs can be seen as equivalent to that between short and long-run measures of comparative advantage. The former will be relevant in the allocation of resources for - 129 - Technical Appendix 4.01 Page 5 of 9 rehabilitation - suwh as general import support - whilst the latter will be appropriate for considering the allocation of funds for investment expansion. 12. Short-run DRCs locus on the costs of generating or saving foreign exchange under existing production conditions, which may include low capacity working. machinery that is poorly maintained and scarcities of key inputs. The relevant costs for inclusion in the calculations are variable costs of operations, plus any maintenance costs for capital assets. Full capital costs - defined as the replacement value of investment calculated as an annuity at an opportunity cost rate of discount - will not be relevant since in the short-run these will be sunk costs, that will have to be met regardless of whether production is expanded in the short-term 5/. Strictly therefore the costs relevant for short-run DRC estimates are variable not full costs of production (capital plus variable). 13. A distinction sometimes made in DRC studies is between short-run DRCs, at existing capacity working, and those at full capacity. This is to allow for the fact that constraints on achieving full working may penalize enterprises which would be economically efficient were they able to produce at their design level of output. Clearly the differences between these two versions of the short-run DRC will be greatest in lines of production where economies of scale in operations are most significant. However, given the difficulty of establishing what costs would be like under the hypothetical conditions of full capacity working this second version of the DRC measure is not applied in the present study. 14. The long-run DRC indicator, assesses the full cost of expanding production to generate or save additional foreign exchange. As it focusses on the longer-term it covers situations where new investment is carried out to expand capacity to produce traded goods. Here capital costs will not be sunk but must be included in the calculat-ons annualized at an appropriate discount rate and expressed per unit of output. Data on investment costs for different sub-sectors of manufacturing were difficult to obtain, so that the DRC estimates given here focus primarily on the short-term. Long-run DRC calculations that build on the work of this study should be part of follow up work on industry in Mozambique. DRC estimtes 15. The figures refer to operations in 1987, except in the case of sugar where the average world price of 1988 - 89 is used, which is 502 5/ The only capital cost that is relevant is where assets have a scrap or resale value. Our estimates have ignored this factor since in existing conditions there is unlikely to be a significant market for second-hand assets _ 130 - technical Loewndix 4.01 Page 6of 9 above that of 1987. All costs are converted into dollars at the prevailing official exchange rate. DRC ratios are estimated using equation (1), except for one modification. Electricity costs are broken down Into 502 non-traded costs and 502 traded costs. The non-traded Electricity costs therefore are included in the numeratcr of the ratio, whilst the traied costs are in the denominator with a negative sign. Short-run DRCs are estimated and there is no allowance for a capital charge. 16. Utilizing these data short-run DRC estimates are available for 2 sugar mills, 2 tlmber mills, and the cashew producer. The results are as follows$ Cashew 0.57 Sugar Mlll 1. 1.50 Sugar Mill 2. 2.00 Timber Mill 1. 1.1? Timber Mill 2. 0.58 17. They indicate results which are intuitLvely not surprising. - shew production ls competitive - thls is clearly not unexpected, since it a major export industry. Sugar is high cost with both of the mills owing high DRCs, although mlll 1 may be marginal at a shadow exchange te. The position in relation to Timber is ambiguous since the smaller 11 is efficient, whilst the latter has a DRC of above 1...However, ain, at a shadow exchange rate more than 202 above the official rate thls c!ond mll to be classed as competitive in short-run terms. - 131 - Technical Appendix 4.01 Page 7 of 9 DRC calculations for Agro-Industrial Activities 198' (units US$ thousands at 1987 prices) Cashew (Caju de Mozambique) Foreign Domestic Exchange Resource Costs Output 29,384.0 Inputs Inputs Electricity 145.5 Electricity 145.5 Services 33.3 Services 937.6 Raw Materials 0 Raw Materials 14,223.8 Salaries & Wages Salaries & Wages 878.0 Other 392.3 Net Foreign Exchange Effect 29,174.2 16,577.2 DRC - 16,577.2 - 0.57 29,174.2 Sugar 1. (Sociedade Agricola Incomati) Foreign Domestic Exchange Resource Costs Output 7,326.6 inputs Inputs Electricity 890.0 Electricity 890.0 Services - Services 92.5 Raw Materials - Raw Materials 4,215.0 Salaries & Wages - Salaries & Wages 2,850.0 Others 1,810.0 Net Foreign Exchange Effect 6,436.6 9.857.5 DRC - 9,857.5 - 1.53 6,436.6 - 132 - Technical Appendix 4.01 Page 8 of 9 Sugar 2. (Acucareira de Mapambisse) Foreign Domestic Exchange Resource Costs Output 4,398.0 Inputs Inputs Electricity 642.5 Electricity 642.5 Services - Services 1,580.0 Raw Materials - Raw Materials 2,815.5 Salaries & Wages - Salaries & Wages 2,500.0 Others 100.0 Net Foreign Exchange Effect 3,755.5 7,638.0 DRC = 7638.0 - 2.03 3755.5 Timber 1. (Madeiras de Cabo Delgado) Foreign Domestic Exchange Resource Costs Output 1,753.1 Inputs Inputs Electricity 56.5 Electricity 56.5 Services 359.8 Services 885.7 Raw Materials - Raw Materials 241.4 Salaries & Wages - Salaries & Wages 199.0 Others 182.5 Net Foreign Exchange Effect 1,336.8 1,565.1 DRC = 1565.1 - 1.17 1336.8 - 133 - Technical Appendix 4.01 Page 9 of 9 Timber 2. (Ifl.ma) Foreign Domestic Exchange Resource Costs Output 5,411.4 Inputs Input Electricity 194.9 Electricity 194.9 Services 401.3 Services 624.2 Raw Materials 642.5 Raw Materials 566.5 Salaries & Wages 291.3 Salaries & Wages 375.8 Others 490.9 Net Foreign Exchange Effect 3,881.4 2,252.3 DIC - 2252.3 - 0.58 3881.4 Notess 1. Data supplied by the Ministry of Agriculture. 2. *. 1 values are in 1987 prices, except for sugar (1988). 3. Electricity costs are divided into 5OX traded and 50Z non-traded. according to cost estimates supplied by Electromoc. 4. The conversion factor used to compare domestic and world prices for sugar is particularly uncertain, given doubts about the accuracy of the cif price of sugar to Mozambique. 5. Inclusion of financial charges of depreciation and interest - as a proxy for capital costs do not alter the results significantly for Cashew, and Sugar. However, the DRC for Timber 1 rises to 0.96 and for Timber 2 to 1.44, with the inclusion of financial charges. _ 134 - Technical Appendis 4.02 Page 1 of 11 ANALYSIS O EFFICIENCY OF INDUSTRIAL SUBSECTORS. 1. This appendix provides details of the subsectors analyzed under the DRC ratio study of Chapter 4. SIC 31 - Agroprocessing: (5 products/4 enterprises) 2. The two sugar mills in the sample had estimated short run DRC ratios respectively of approximately 1.5 and 2.0, indicating inefficiency vis-a-vis average 1989 world sugar prices of US$0.23/kg). The primary causes, low yields. low refinery utilization capacities, high labor overheads and security problems are well known. The Government is planning case by case studies to develop rehabilitation (or abandonment) strategies. In the case of Mapambisse, significant production cost savings would apparently be needed for long run viability to be restored, or alternatively a significant additional real increase in the world price of sugar. Feasible exchange rate devaluation alone would not be able to restore economic viability. 3. Caju de Mozambique (DRC ratio 0.6) proved to be an efficient generator of foreign exchange in the short run. This is to be expected since up to 1975 the industry was a major world exporter. 1/ 4. Based upon 1987 cost data from a private sector refinery, cottonseed oil refining was clearly inefficient. However, the DRC. ratio is uncertain for lack of reliable information. Purchase prices for cotton seed were $100/ton at the official exchange rate and $163/ton in warehouse. Using a conversion ratio of 7.69 tons/seed per ton of crude oil (13Z oil content vs. maximum of 162), the unit cost of one ton of local crude oil is $1325/ton whereas the CIF Rotterdam prices for the period were in the range of $550/ton. At this level of crude oil cost, the cost of refined oil ($1,616/ton) would be clearly uneconomic. 2/ Analysis of the precise causes of the refining cost inefficiency requires investigation at the enterprise level, particularly concerning the conversion ratios from crude. Rehabilitation of a similar refinery, FASOL, is now being studied 3/. Erratic supply of domestic raw material and subsequent low utilization rates are characteristic of edible oil refining in recent years. 5. Extra laundry soap, produced by the same enteriprise from local coconut oil and imported soap tallow had a DRC ratio of below 1.0, when 1/ The other principals are Tanzania, India and Brazil. 2/ Part of the edible oil consumption of the Maputo area market is currently being supplied through donor program imports of crude oil which is refined and packaged locally. 31 Under the IDA Industrial Enterprise Restructuring Project. - 135 - Technical Appendix 4.02 Page 2 of 11 compared with imported CLARIN laundry soap, implying efficiency in the short run. Higher quality soaps were not as competitive however. SIC 31 - Food Processing (3 products/2 enterprises) 6. DRC analysis in the foods subsector is rendered difficult by the high incidence of de facto non-tradeables due to the natural protection of distance, packaging and perishability problems; difficulty of product comparability; and product multiplicity in individual enterprises. 7. Spaghetti (unseasoned without preservatives) from two enterprises had short run DRC ratios of 0.6 and 0.5 respectively when compared with imported spaghetti at the duty free store, even after allowance for assumed 1002 markups over CIF prices. Cream crackers had a DRC ratio of 0.5. Even allowing for the imperfections of the analysis, the results indicate a reasonable level of efficiency for these relatively simple, mechanized processes which use old, but functional and fully amortized equipment, and whose labor costs are only 62 of local processing costs. SIC 32 - Textiles and Clothing S. The cotton textile and garment subsectors already have the elements in place for a fully integrated industry. Three cotton textile mills (TEXMOQUE, TEXTAFRICA, TEXLOM) have a combined design production capacity of 47 million meters of cloth annually; and some 15 ready-made garment factories are located in Maputo and Beira. Estimated total production of the three textile mills in 1988 was only 21 million mt or 452 of capacity utilization. Since 1986, domestic cotton supply has not been sufficient to meet even this demand. It is considered however that under secure rural conditions and a rationalized pricing structure, cotton supply can expand rapidly to meet local demand for both textiles and edible oils. 9. A synthetic fiber mill produces TREVIRA (70. polyester/30Z viscose) fiber of reportedly good quality, with 3.9 M Mm2/yr. rated capacity. It intends to export although it is a relatively high-cost producer, with significant expatriate staff costs, and relatively low capacity utilization. Average quality TREVIRA material is priced ex factory at $4.75/metre, vis a vis HongKong source equivalent quoted at approximately US$1.17/yard CIF Maputo ($2.40/yard after all duties and taxes). The factory gate cost of the mill's raw material (dyed and blended fibre) in 1987 was $9.50/kg (source from Portugal), apparently well in excess of the world market price. Input price levels, high technical assistance and finance costs, and protective tariffs render the overall operation clearly cost inefficient; however the - 136 - Technical Appendix 4.02 Page 3 of 11 mechanical efficiency of the weaving operation itself is reported to be satisfactory. 4/ 10. The domestic cotton textile industry is currently a high-cost producer by world standards. Recent invoices provided by management indicated that Hong Kong source cotton poplin has a CIF Maputo price of around US$0.90/yd. (and after 76Z total fiscal charges, a factory door cost of $1.59/yard). Equivalent domestically produced material had an ex-factory price of $1.75/yard. Il. Partial rehabilitation of the Textafrica mill was financed by the First World Bank Rehabilitation Credit, and the Texlom mill by French assistance. The French consultants consider Texlom to be potentially competitive if permitted to operate on strictly commercial criteria, a proviso which would require shedding of much employment. Current production costs are estimated to be more than twice as expensive as those of Europe for equivalent specifications. 12. At present, the cost structure of the Mozambican textile industry does not permit international cost competitiveness for garments manufactured using locally source textiles. Achieving competitiveness will be a medium to long term project requiring extensive rehabilitation of both cotton production and mills. 13. In the short term, however, it may be possible for Mozambique to become a garment exporter based on imported fabric, by taking advantage of the reduced industrial labor costs, and ample supply of low cost factory space available in Maputo and Beira. A possible indicator of competitiveness of labor costs is that SOVESTE E.E., a garment producer, has established a joint venture with a Hong Kong firm to produce shirts and trousers for export to Europe from imported material 5/, and a Bulgarian state firm to produce workclothes from imported material for export to Bulgaria. Garment export may provide a commercial opportunity, providing a short-term source of employment generation and foreign exchange, while the longer-term project of rationalizing the domestic textile industry is being carried out. 4/ This case illustrates that price non-competitiveness can be generated from a variety of sources quite independent of plant technical operating efficiency, thus highlighting the fact that neither policy recommendations, nor enterprise rehabilitation schemes can be formulated intelligently unless the specific sources of inefficiency are understood. This latter almost always requires individual in-depth enterprise analysis. 51 The Hong Kong firm supplies key technical personnel, and SOVESTE the labor force, factory building and equipment. _ 137 - Technical Appendiz 4.02 Pa e 4 of 11 SIC 32 - Leather and Shoes (3 enterprises) 14. Included in the sample were two manufacturers of canvas tennis shoes, and IMOCAL, a state enterprise, which directly manages five leather shoe factories with 890 employees. 15. Accurate DRC ratios for shoes are difficult to estimate due to the product comparability problem. Estimates for one of the factories which produces low quality canvas shoes with obsolete equipment indicated that DRC ratios were less than 1 (efficient production). Local manufacturers indicated a consumer preference for significantly higher priced but higher quality imported products when available. 61 16. Leather shoe production in Mozambique is complicated at present by a shortage of domestic leather due to the security situation. About one half of leather used for shoe production is imported. Comparative price information supplied indicates that domestically produced leather and shoes are cheaper generally than imports considered to be equivalents. An indication that shoe manufacturing in Maputo may be efficient is that a South African manufacturer has been negotiating the lease of one factory with its work force to produce for export. The machinery, equipment and production supervision would be supplied by the South African firm. SIC 33 - Wood and Wood Products (2 enterprises) 17. A number of small furniture making enterprises and a saw mill were interviewed in addition to the two enterprises for which DRC ratios were estimated. The nature of the product lines and accounting of these small enterprises is such that reliable DRC estimates could not be attempted. However one small saw mill operation in Beira had secured an export order for tropical hardwoods t * loilth Africa at a contract price of $35S0m3 FOB Beira. The mill estimate* production costs at approximately $310/m after taking account wastage of 502, a rate acceptable for an operation of this nature. Furniture making operations use attractive domestic hardwoods, and semi-artisanal production methods. Production can be assumed to be virtually non-tradeable, but quality is satisfactory by international standards. 18. The two forest complex operations for which estimates were compiled had short run DRC ratios of 1.2 and 0.6. The first enterprise would be efficient on the assumption that the real exchange rate was over 202 above the official rate. However, it would probably be inefficient in the long run after the inclusion of capital costs on present productivity levels. Both have been seriously affected by the security situation. The latter is 6/ One of the tennis shoe producers is now being studied in detail within the context of the IDA Industrial Enterprise Restructuring Project. - 138 - Technical Appendix 4.02 Page 5 of 11 operating at only about 30-352 of capacity utilization. 7/ It is difficult to draw firm conclusions based upon these estimated DRC ratios, given the current security and data problems. Under normal security conditions, Mozambique has in the past been a competitive exporter of senm timber. SIC 34 - Paper and Paper Products (3 enterprises) 19. The enterprise for which ratios were estimated produces cardboard from imported kraft paper, and cartons. A second firm buys cardboard from the first and also producers boxes. These industries are critical to local export and import substitute producers, and are relatively cost efficient, permitting flexibility of production to meet special local packaging specifications. Estimates show DRC ratios of below 1.0 indicating efficiency. A third enterprise recycles locally purchased waste paper to produce notebook paper. Design capacity is 6000 tons/yr. but production is restricted to 3000 tons/yr. by lack of locally available waste paper. The enterprise itself is now doing a cost/benefit study comparing direct importation versus local production as part of a feasibility study for a proposed Swedish financed $3 million plant rehabilitation project. In the absence of cost data, it is probable that a plant recycling local waste material. with plant and equipment treated as a sunk cost, would be an efficient saver of foreign exchange in the short run. However, in the longer term the economics of the process at low volume may be marginal. The justification for a long term investment is under study. SIC 35 - Chemicals, Rubber and Plastics (5 enterprises) 20. This subsector includes a diverse variety of producers in terms of both enterprise size and product lines. Included in the sample were a tire manufacturer, two producers (comprising six factories) of moulded, extruded and film plastic products, a paint manufacturer and a producer of paraffin candles and vinegar from industrial acetic acid. 21. The tire factory was the only enterprise in the sample for which sufficient data was available to calculate DRC ratios at both the existing utilization level (approx. 331), and full utilization respectively for 1100 x 20 SE truck tires. In both cases, results showed that production was efficient. This plant was established in 1979 and pays royalties and fees for production technology and expertise to an international producer. The enterprise has exported in the past, and the relatively low short run DRC ratio at full capacity utilization implies that economic viability at full cost may be attainable. 22. The candle and vinegar producer is a small enterprise which imports raw materials (acetic acid and paraffin) from South Africa despite 7/ One enterprise, Madeiras de Cabo Delgado, is currently being studied in detail under the IDA "Industrial Enterprise Restructuring Project'. - 139 - Technical A.pendix 4.02 Page 6 of 11 a lack of ready access to foreign exchange through official channels. (a common problem for smali producers up until the startup of the SME fund - see chapter 5) 8/. At the official exchange rate the DRC ratio was 0.8. Evaluating domestic output at the parallel market rate at which imported inputs were obtained (MT 1300/lt), the DRC ratio was 0.3. The ratio for vinegar production could not be directly estimated, due to the lack of cost breakdown for containers which consist of recycled bottles purchased from 1eual refuse collectors. The operation is clearly efficient on the base of price competitiveness with duty-free imports, and natural protection because of the weight gaining nature of the industry (942 by volume of the end- product being water). At least six small-volume producers are active in Maputo. 23. The plastics extrusion and moulding industry is characterized worldwide by a multiplicity of producers distributing for local markets. Eight producers in Mozambique will produce an estimated 1500 tons of product in 1989. Four items were evaluated. In the absence of usable price comparisons, the table shows the CIF price below which local production would be efficient (the price for which the DRC ratio - 1.0). Product CIF Maputo at DRC - 1 110 nm (41) Duronil PVC pipe (l meter) $2.53 314 inch PVC Garden Hose (1 meter) $0.39 2-112 liter Polyethylene Plastic Bottle $0.40 Size 38 Transparent Ladies Sandal $0.92 24. In lieu of DRC ratios, there are indications that these items are reasonably price competitive, since locally produced PVC pipe is reportedly price competitive with imports for bids for the Beira Corridor project. In addition, there is an element of natural protection for local production of high volume configurations such as bottles and large diameter pipe. SIC 35 - Paints 25. The paint industry worldwide is characterized by many relatively small production units operating in relatively restricted geographical markets. The paint industry in Mozambique consists of 4 operations, one in Beira, and the rest in Maputo. All import raw materials in bulk, then bland, shade, mix and fill locally. Cans are sometimes provided by Metal Box and sometimes imported. Finished production is sold entirely in the domestic market. Capacity utilization rates have typically been low (lOZ-15) due to a lack of foreign exchange for raw material import. The low utilization rates in the existing plants do not seriously affect direct production costs 8/ The paraffin candles prices were compared with equivalents available at the duty free shop in Maputo, assuming 10OZ mark- up between CIF and shelf price. - 140 - Technical Appendix 4.02 Page 7 of 11 due to the relatively unsophisticated (and fully amortized) equipment and batch processing operatior. The advantages of local production, in spite of the relatively high import content (70?X), are the ability to adjust formulations and packaging requirements to specific orders. 26. Due to the multiple product lines and formulations characteristic of the paint industry, direct comparators were not available for DRC calculation. Nevertheless, production cost data supplied by a local enterprise and a sampling of import price quotations for paint, permitted some inferencer to be made for one enterprise. Excluding solvents and primers, unit production costs ranged from $2.77/lt to $3.49/lt. According to price comparisons for imported paints (obtained from UCPI), the prices of a typical white paint 'Extra White* fell to below CIF price (S. Africa origin) after the January 1987 devaluation from a level significantly above prior to the ERP. Quotations for May 19 8 FOB France, were as follows: Local Enterprise Product Unit Cost $US/l. Approximate Comparator Cost FOB France Ce..luc Brilliant 5.75 4.50 Celluc Brilliant Rouge 5.74 4.50 Alum. Redox Rui 2105 14.61 5.09 Diluent for Rui 2105 3.22 1.53 27. While the comparators are not exact equivalent products, the enterprise production cost figures tend to indicate that for those formulations which can be successfully produced in Mozambique, production and sales costs are competitive and in DRC ratio terms, the local processing is probably efficient, particularly for packaged rather than bulk products. SIC 36 - Non-metallic Minerals 28. Cement: Cimentos de Mozambique has three integrated production units located respectively in Nacala, Dondo (near Beira) and Maputo. Total design capacity of the three plants at time of construction was 970,000 tonslclinker/yr. In 1988 Nacala produced 23,000 tons of clinker and 24,000 tons of bagged cement. The clinker units at Maputo have been shutdown since 1984 due to insecurity of raw material transport. The Dondo clinker unit has also been shutdown for several years as limestone supply for Dondo is over an as yet unsecured railroad line. In 1988 total national production was - 141 - Tecbnical Appendiz 4.02 Page 8 of 11 69,000 tons. The supply deficit is made up with imports from a variety of African and other souirces 9/. 29. For the current grinding and bagging operation using local clinker, sbort run DRC ratio would be below 1.0 at a cement price for bagged cement of $65/ton. (Quotations for imported bagged cement CIF Maputo or Beira were in the range of $S0-70/ton). This operation, in the short run is therefore efficient as an import substituter. However, with imported clinker it is not efficient. There is a benefit associated with local production due to the lower breakage rate which is a problem with imported bags. Additionally, in the absence of local production, ability to negotiate competitive prices with foreign suppliers is often more difficult. In the lorg run the economics of cement production depend critically on the transport costs of materials and products. With rehabilitation and integration of production, if operated at reasonable capacity utilization levels, the costs of production could be radically changed. Any conclusions require specific studies, which are in process. 30. Glass: Vidreira, SA, located in Maputo. is the only producer of glass and glassware in Mozambique. The operation currently has a 72 tons/day fuel oil fired furnace for bottle production and two 12 ton/day electric furnaces for production of domestic glassware. Raw material supply sources are local (sand, limestone) and foreign (feldspar, sodas). Factory door price of the imports is 302-40Z higher than CIF after taxes and duties. Operation in recent years has been sporadic due to frequent power outages and lack of back-up power supply. 10/ The plant currently has 1000 employees, but it would be possible reportedly to reduce this to 800. 31. Unit operating costs in recent years have reportedly been more than twice international costs due to low capacity utilization and the frequent shutdowns, which necessitate costly furnace restarts. Management states that current marginal costs are about $0.70kg for domestic glassware, roughly equal to world costs. Quality problems in bottles have occurred due to quality of sand supply. Export potential exists in neighboring countries for certain lines (kerosene lamps, cups, glass). The plant maintains only general accounting, preventing reliable DRC analysis. It would appear that the plant is marginally efficient in terms of operating costs. Information was not available to make any inferences concerning relative efficiencies after the planned rehabilitation. 9/ Rehabilitation of the Matola clinker unit to a capacity level of 450,000 tons/yr is now being financed by NORAD, and the CEE. A pre-feasibility study of various options for Dondo is being carried out within the context of the Industrial Enterprise Restructuring Project. 10/ Italian grant financing is proposed to reconstruct the plant and facilities, including provision of backup power. - 142 - technical Appendix 4.02 Page 9 of 11 SIC 38 - Metal Products and hachinery (14 enterprises) 32. The subsector can be classified into those operations producing either industrial or consumer goods on an assembly line basis; and fabricating shops producing customized pieces to order. Among the former group included in the sample were producers of electric transmission cable; bicycles; replacement auto radiators; metal containers and cans, agricultural implements; refrigera..ors, sewing machines, horizontal freezers; metal furniture; well puimps; and air conditioner assembly. Among the fabricating shops were one which fabricates large structures and railroad cars; and several others which fabricate items such as tanks, tractor wagons, truck bodies. etc. and metal hospital furniture. 33. For the former group, is was possible in some cases to calculate DRC ratios. In the case of fabricating shops, it is generally not possible due to the specific nature of each order. Significant foreign exchange is saved through local fabrication due to saving in transport costs of bulky fabricated items such as truck bodies, water tanks, etc. Moreover, local quality is in most cases relatively good, although little design innovation has taken place since Independence. A summary of results for the engineering sector is as follows: Product Comments Electric Transmission Per available data, DRC ratios of 0.3 for four cable specifications. Underestimate possible due to under valuation of raw material. Parent company to establish joint venture for export as well as domestic sales. Bicycles (one-speed DRC ratio 0.4 vs. Zimbabwean equivalent (CIF utility)price $90). Frames and wheels are fabricated and painted locally; other components are imported. in kits for assembly; enterprise attempts to keep production costs in vicinity of $50/unit and after tax sales price e $70. Import restrictions effectively protect vs. better quality Asian and Indian bikes in similar price range. Security situation has seriously affected sales to principal rural markets. Temporary shut-down in early 1989 due to inventory build-up. Replacement auto DRC ratio 0.3. Efficient foreign exchange radiators saver for replacement items if cost data provided is realistic. - 143 - Technical Appendix 4.02 Page 10 of 11 Metal containers Packaging industry hampered by obsolete equipment and lack of direct access to foreign exchange for materials. DRC ratio estimate not possible for lack of comparator Imports. Effective utilization rate in 1988: 252. Market reduced by installation of canning by Cashew processing factories. Agricultural implements A technically competent producer but security (ploughs, disks) situation has reduced market resulting in inventory build- up. Is reportedly price competitive in SADCC market (e.g. disk plough price $840 vs. ex-works Harare of $1,400 for equivalent). Natural protection--freight is 25t cf cost of imports from Europe. Quality is competitive. Refrigerators, DRC ratio of 0.5 vs. South African equivalent horizontal for refrigerator CIF Maputo Rs. 1500 (two- freezers, mechanical door 11 ft.3 unit). Price competitive with South and sewing machine. Africa for mechanical, but not electric, sewing machines. Considerable natural protection afforded via transport cost of imported equivalents. Maximum CIF Maputo prices for equivalents at DRC - 1, are as followst 9 ft3 refrigerator - $351; 11 ft3 refrig. - $452. Sewing machines: mechanical - $107; electrical - $213. 240 lt horizontal freezer - $400. Mechanical Rural Water CIF Maputo for Indian Equivalent $440. DRC Pumps ratio 0.3 for local fabrication of Indian design pump. Raw material cost is only CIF plus 7.5Z due to exemption for rehab. project. Hot Water Heaters (40 1. Same fabricator as water pumps. Probably to 120 1.) efficient due to natural protection (volume) and apparent technical efficiency of fabricator. 34. Metal Fabrication: Aggregate cost data from a leading metal fabricating enterprise (producing truck bodies and waterstorage tanks), indicate the comparative advantage of these fabricating operations relative to other industries in terms of local value added and labor intensity. The import content for the representative fabricating shop is relatively low compared to the industry average. The corresponding figures were 35Z import content, labor 312 of total costs and 482 of local non-tradeable inputs. _ 144 - Technical Appendix 4.02 Page Iof 11 35. Although there is presently little more than occasional export potential to surrounding countries for the metal fabricating subsector, its relatively high labor intensity, and contribution to domestic value added would generate substantial foreign excharge savings when demand for such goods begiDs to accelerate. There exists a pool of fabricating and mechanical skills; however, educational levels, even of skilled operatives, are generally low. and industrial innovation has stagnated for some years. Considerable upgrading and supplementary training of this work force in new technologies which will be required in order to realize the potential benefits. Table 1s MOZAMBIQUE - COST STRUCTURE - LOCALLY PRODUCEI PAINTS Raw Direct Material Labor & Imported Sales Costs (CIF) Customs Factory Production Content Price Product Line Maputo) Duties Expenses Cost (1) (Ex Tax) 1. Latex Enmlstons (house paint) $1.80 $0.54 $0.24 $2.58 70Z $3.63 2. Metallic Paints (Aluminum) 2.43 0.73 0.24 3.40 71 5.09 3. Industrial Paints (enamels) 2.50 0.75 0.24 3.49 72 4.50 4. Alloyed based paints/ lacquer 1.95 0.58 0.24 2.77 70 5.09 5. Primers/ bases 1.70 0.51 0.24 2.45 69 3.96 6. Solvents and thinners 0.80 0.24 0.24 1.28 62 1.53 7. Maritime Paints 2.16 0.65 0.24 3.06 71 *'.22 Overall Average 1.64 0.49 0.24 2.37 69 3.60 - 143 - tecbnical Arnendi. 5.01 Pagel of 3 CALCULAtION OF EPFECTIVE PROTSECTIO RATES. 1. The implications of different nominal tariff rates for effective protection can be illustrated with a simple numerical example. This assumes one output (l) using only one traded input t3), with j 502 of the value of i, when both are valued at world prices. In this case the mPR for i, which measures the extent to which value-added In I at domestic prices exceeds what it would be in the absence of any tariff protection, can be found from a simple formula (tl - *1 t1 ) PRi C 1 - aji ) Z where ti and tj are the import tariffs on the output i and output i respectively, and aji the share of j in the output value of i at world prices. 2. Table 1 shows how different nominal rates of protection on i create different effective rates. Case 1 reflects a situation where all goods have the same lmport tariff - a basic rate of 302. Here the EPR is equal for all activities and is the same as the nominal tariffs of 302. Case 2, 3 and 4 illustrate the effect of giving higher protection to output i, whilst keeping the same protection on input j. Here the DPR rises above the nominal rate. These cases reflect the situation of producers who are to receive special nominal protection, either on infant industry or employment grounds. In the study it is suggested their nominal output protection can be set at the standard rate of 302, plus margins of 102, 202 and 302 alternatively. These additional protection rates will grant a significant increase in EPR since the EPRs in cases 2, 3 and 4, are 502, 702 and 902 respectively in comparison with the standard rate of 302, which is that of those producers who do not receive special favorable treatment. This higher DPR will thus create an incentive for resources to move into these specially favored activities. 3. Domestic producers of luxury consumer goods - that would be subject to an import tariff of 602 under the scheme proposed here - would receive a high EPR of 502 (as in Case 2), if there is no compensating domestic excise tax on the production of these goods. If their local production is to be discouraged on equity grounds there will need to be an equivalent tax that raises revenue equal to 402 of the import price (since 402 is the additional luxury import tariff on these goods). With this domestic tax the EPR on their production will fall to the standard rate of 202, so that they will receive no special encouragement. - 146 - Technical Agpendiz 5.01 Page 2 of 3 Table 1s MOWZEBIQUE - ILLUSTRATIONS OF NOMINAL PROTECTION AND EFFECTIVE PROTECTION RATES. Nominal Import Tariff Effective Rate of Protection Protection Case 1 302 uniform tariff on EPR- 0.30 - 0.5 (0.30) Standard Protection output i and input j (1 - 0.5) EPR - 302 Case 2 40? tariff on output i EPR- 0.40 - 0.5 (0.30) Special protection for and 302 tariff on input j (1 - 0.5) Favored Industries EPR - 502 Case 3 502 tariff on output i EPR- 0.50 - 0.5 (0.30) Special Protection for and 30? tariff on input j (1 - 0.5) Favored Industries EPR - 70? Case 4 602 tariff on output i EPR- 0.60 - 0.5 (0.30) Special Protection for and 30? tariff on input j (1 - 0.5) Favored Industries EPR - 90? Case 5 502 tariff on output i EPR- 0.50 - 0.5 (0) Special Protection and zero tariff on input 3 (1 - 0.5) tariff free inputs EPR - 100? Case 6 502 tariff on output i EPR- 0.50 - 0.5 (0.40) Special Protection and 40? tariff on input j (1 - 0.5) high tariff on inputs EPR - 60? Case 7 zero tariff on output i EPR- 0 - 0.5 (0.30) Negative Protection and 30? tariff on input j (l - 0.5) for Exporter EPR - -30? 4. In practice, there will be several complications not accounted for in these examples. Tariff exemptions for particular items will upset the results, making it difficult to establish anything other than broad uniformity between activities. In addition, intermediate inputs may be protected, so that any industries using these inputs will have their EPR reduced, relative to what -147 - Technical-AuDendix 5.01 Page 3 of 3 it would be if these inputs were simply subject to the general protective rate of 302. Also where some quantitative import controls remain these will have *implicit tariff' effects, which in theory should be incorporated in EPR estimates. S. Case 6 shows the effect of an additional import tariff on input j on the EPR of i. Here j itself is subject to an additional protective tariff, so that it has a total import tariff of 40?. With i having a tariff of 50Z, the EPR is 602. This must be compared with an EPR of 70Z in case 3, where ti is again 502. but only the standard tariff of 302 applies to input J. Any extra protection on locally produced inputs thus lowers thp. protection granted to users of those inputs. 6. However, where users receive inputs duty-free, whilst their output is itself protected, the EPR is raised. This is shown in case 5, where i pays no import tariff. Here the EPR rises to 1002, compared with 702 in case 3, and 602 in case 6. This increase in effective protection, when inputs are allowed duty-free is the reason why the study recommends that tariff exemptions on capital goods and raw materials be withdrawn for all but export producers. 7. The case of exporters differs from that of produce -selling in the local market, since unless exporters are subsidized they will receive only the world price for their output. This is eq"ivalent to zero tariff protection. However if exporters must themselves pay tariffs on the inputs they use, they will be penalized by receiving neEitive protection from the tariff system. This is illustrated in case 7, where ti is zero, but the standard import tariff of 302 applies to input J. Here the EPR is -302. This implies that value-added is 302 below what it would be with no tariff protection, so that the producer concerned is losing from the tariff system. If as is recommended here exporters pay no import tariffs or have their tariffs refunded through a drawback scheme then the EPR in case 7 will be zero. There will remain some bias against exporters, where the standard EPR is 302, but under the proposed arrangements there should be no~ negative EPR. -148 - - - Tecbnical AIpendix 5.02 Page 1 of 3 EFFECTS OF CHANGES IN TAX INCENTIVES 1. The possible impact of the changes in incentives discussed in thc vext can be illustrated with a numerical example. Table TA 1 sets out data on new investment project with a cost of US$1.5 million (at the official exchange ite), and a pre-tax profit of US$0.65 million in a normal year of operations. we project has a 15 year life, with production starting in the year after the _ vestment expenditure has been made. For simplicity assets are depreciated over 15 years and all profits after depreciation are distributed as dividends. Revenue and all costs are constant over the life of the project. 2. Table TA 2 shows taxes and net project income over time under different tax-incentive packages, whilst table 3 summarizes the different cases and the total taxes collected. Case 1 illustrates a situation with no special tax concessions. Profits and Dividend Taxes are at previous rates of 502 and 55Z respectively, and there is no special depreciation allowance. The project has an IRR of 142 post-tax and US$7.56 million are collected in tax revenue. Case 2 introduces a 5 year tax holiday during which no profits tax is paid, and where Dividends Tax is removed. During the first five years (1-5) there is a double depreciation allowance. Here the project IRR rises to 482 and total taxes collected fall to US$2.75 million. 3. Case 2 can be taken to be a fairly representative situation under the existing incentives. Case 3 illustrates the most favorable incentives possible for a new investment under the existing legislation since the tax holiday period extends to 10 years. this is also a case in which the double depreciation allowance is redundant. Here the project IRn rises to 49Z, whilst taxes collected are US$1.63 million, about 202 of those in case 1. 4. Cases 4, 5 and 6 show the effect of the changes proposed in this study. In all three cases there is a 202 import tariff on the capital equipment used by the project which is assumed to be all imported. In case 4 with a basic rate of profits tax of 50S there is a two year tax holiday. In case 5 this tax holiday is removed, but the basic rate of tax is reduced to 402. Case 6 shows the effects of a 502 rate of profits tax and no tax holiday. 5. In these cases tax revenue is considerably higher than in case 2 and 3. This is true both in terms of absolute tax receipts and in terms of present values. Lower basic profits tax instead of tax holidays in initial years increases the present value of tax revenue for a given absolute amount. For example in case 5 with lower basic tax rate of 402 the present value of tax revenue is double that of case 2, and nearly 4 times that of case 3. 6. Although these figures are no more than illustrative they show potentially large gains in tax revenue from restricting tax exemptions and holidays. Such gains can be used to offset the loss in revenue from removal of dividends tax and future reductions in profits tax. Changes in these taxes also affect all producers, not only new investors, so that the concessions are spread more widely. 149 Technical ADDrndix S.02 Poge 2 of 3 Table 1s MOZAHBIQUE - YPOTBL'TICAL PROJECT Investment Cost US$ mill. Equipment1/ 0.75 Buildings 0 .75 Sales Revenue bl 1.50 Operating Costs Raw Materials 0.50 Labor 0.25 DepreciationCl 0.10 Profits pre-tax 0.65 Notess a/ All Imported bI In normal year of operations cl Over 1S years. Table 2: MOZAMBIQUE - TAX INCENTIVE PACKAGE AND TOTAL TAXES COLLECTED Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Profits Tax 50Z 50Z 502 S02 402 5oz Tax Holidays Zero 5 years 10 years 2 years zero zero (Profits) Import Tariffs None None None None 202 202 on Capital Equipment Depreciation Normal Double Normal Normal Normal Normal allowance 15 yrs 5 yrs 15 yrs 15 yrs 15 yrs 15 yrs a/. Total Taxes 7.56 2.75 1.63 4.37 3.90 5.02 collected (US$ mill) Discounted value of Taxes 4.33 1.18 0.80 2.35 2.22 2.93 collected (at 82) (USS Mill.) Project InR (1) 14 48 49 34 29 25 al Double depreciation allowance carried over beyond year 5 when profits tax is due. - 150 - Technical Appendix 5.02 Page 3 of 3 7. The incentive package in cases 4, 5 and 6 clearly reduces the project IRR below that in case 2 and 3. However this is only a problem if the IRR after tax is insufficient to justify the investment so that firms will not choose to invest under these rates of tax and projected profits. Provided the project IRR is above the minimum necessary to stimulate investment, the fact that it is less than could be obtained under an alternative incentive package should not be an issue. 8. The labor subsidy cas credit referred to in the text can be illustrated with the same project example. It is assumed that on average workers are paid twice the minimum wage, so that labor costs at the minimum wage are US$0.125 million. The value per worker for calculating the credit is taken to be 30OZ of the minimum wage. Therefore the total tax credit that is deductible from Profits Tax liability is US$0.375 million. 9. In a normal year of operations taxable profits are US$0.65 million, so that at a 402 rate of tax, Profits Tax is US$0.26 million. With a tax credit of US$0.375 million taxable profit falls to US$0.275 million and Profits Tax collected falls to US$0.11 million. However in this particular case, where labor costs are low relative to profits, this form of tax credit is not a major incentive. With a Profits Tax holiday of 2 years, for example, the firm will save a tax bill of US$0.26 million for two years. However with this tax credit the saving will be US$0.15 million per year. If the project were to employ twice the number of workers at the same wage rate its credit under this scheme would double, with a net saving of US$0.30 million per year. If the tax credit is transferable between years this would be a more favorable incentive to the firm than a straight two year tax holiday. The degree to which firms would gain from this type of incentive thus depends on the extent to which labor is an important element of costs of production, and the level at which the incentive is set, determined by the ratio of the value per worker used to calculate the tax credit and the minimum wage. - 151 - Technical Appendix 6.01 Page I of 1 OPRATING MECHANIS1 FOR REVISED EXPORT RETENTION SCEEME. 1. A possible operating mechanism for a revised export retention system based on special accounts in an overseas bank would be as follows. 2. A special account would be organized into a system of sub- for each exporter. The functioning of the special account would be ed by BM and the foreign bank would have a service contract ng how the overall :ccount and sub-accounts would function. 1/ The ncept behind these accounts is to provide automatic, on going access gn exchange based on performance. A contractual agreement 2I between rnment and the exporter would outline the responsibilities and goals t. The retention agreement (as is presently done) would be lased on ___ ign cost percentage of exports. and the exporter would agree to handle all transactions through the special account and receive local currency for these funds from BM at the official rate of exchange. The agreement would also stipulate that the exporter would lose access to foreign currency if targets were not met. 3. All payments to the exporter would have a fixed percentage credited to the sub-account and the balance of funds would te credited to a single Government sub-account available for other industrial sector exports or imports. For example under the account "Banco de Mocambique-Currency Allocation Account, the foreign bank would establish two sub-accounts: Account A- Company Allocation Accounts, and Account B- General Account. Account A would consist of various sub accounts denominated for each Company, while Account B would be a single account and constitute the pool of surplus foreign exchange available to the Government. 4. The system could also allow as an additional incentive to exporters a further set of subaccount based on a fixed percentage formula for free foreign exchange for imports. This could inter alia encourage expansion of exports. 1/ A similar system as proposed herein has been devised by the Swedish Development Authority with the Government of Tanzania, Nicaragua and Viet Nam (under study). These accounts are smaller than what would be the case for Mozambique but discussions with banks revealed that it would be relatively simple to maintain and operate the account as described. 2I This agreement would be identical to the present agreement signed between the exporter, BM and the Ministry of Commerce. - 152 - Technical A2pendiz 6.02 Page I of 2 COMPANY PRIVATIZLTIONs ILLUSTRATION 0? PORK OF MANAGENENS CONTRACT AND COMPANY TRANSFER PROCESS. i) Rights of Management$ The Government would hire the partnership to manage one or more industries under its control with the following conditions: - The Government would commit a specified additional equity investment (voting shares) in the company to be made in a specified time frame. - The Government would further ensure the provision of debt as estimated in the requirements for rehabilitation. - Insofar as possible the Government would also guarantee on-going access to foreign exchange for short term inputs. - The Government would delegate all aspects of management and control to the partnership with the important exception of any sale of assets. - The management contract would cover a period of between 5-10 years. ii) Performance Conditions: The Government would reserve the right to intervene only in the case of malfeasance or if the following performance criteria were not met: _ Six month interim and annual final audit by an independent auditing firm determined by the Government. Sales and profitability falling below the worst case scenario estimated for the industry. 1/ If the lack of performance is due to inability to obtain foreign exchange at the level defined in the worst case scenario the management contract will continue in force. iil) Rights to Purchase and Income: The principal purpose of the agreement is to ensure a timely and proper sale of the industry which the partnership agrees to buy, overtime, at a price determined by: a) Equity Investment b) Annual earning on equity at a rate of _ 2 per annum. Profits of the enterprise incur tax at ---- Z of profits and would be used in the following manner after tax: 1/ As part of the feasibility study and negotiations three hypothetical scenarios should be mapped out for the industry: a best case, worst case and probable case scenario. The purpose of these scenarios would be to find a realistic 'median" for determining price and buy out conditions. - 153 - Technical Aupendix 6.02 Page 2 of 2 102 allocated to bonus fee payable to Partnership (salary and other allowable costs are paid out of overheads of the enterprise). 50t allocated to purchase of shares (calculated by the original value of shares plus annual interest rate capitalized annually). Of these shares 602 allocated to partnership and 402 to a workers participation scheme. 2/ The workers will elect a committee to oversee their investment. 401 allocated to reserves of the enterprise. Until all voting shares are purchased preferred shares will not receive dividends. After the voting shares have been purchased preferred shares will receive a fixed proportion of the rights of ordinary shares. iv) Illustration of process of purchase of companys Assuming the following situation for a company prior to new equity investments Asset value (after revaluation-debt) considering market values S 1,000,000 Profits (before tax): $ 50,000 Government positive net worth: $ 1,000,000 This basic data would show a marginally profitable company with actual net worth and market value of assets being equivalent. Assume that a rehabilitation study shows that with an additional $1,000,000 in investment the enterprise could make a profit of $150,000, on the basis of 502 of the new investment supplied as equity. In this case the above type of management contract would allocate as followst 102 Bonus to Partnership $15,000 502 Purchase $75,000 401 Reserves $60,000 If the equity investment is to earn 10 per annum it would take over 10 years to complete the purchase assuming profits did not grow. If the reserves were used then purchase could be accomplished in 5-7 years. After this period the original investment could receive 1/3 of profits which would be equivalent to its initial return on investment ($50,000). 2/ In principle it would appear best to avoid any possibility of conflict in ownership and control between labor and management. The critical importance of and incentive to management further justifies the 60140 esplit'. - 154- NE HOZAMBIQUE ANX INDUSTRIAL SECTOR STUDY page I of 1 lbwiW: WSuW kects d the ik d Ode n luueto S tbe km St_aid lottb de lzu_, 1wo (In nlillaus of neticus) , ................................. . ..... . 1164 185 186 1988 1588 I lurh June Slot. II:. frch JUN ................................................................................... .................................................................................... Seuervn 7608 1I3 17413 28242 32 3671? ... 446 cub 14 1%36 1688 2068 19 2382 . 3217 Depsits mith 1I/ 56%4 12255 ISmS 27174 30520 34335 .. 1449 feap asuts 55 55 6IC 5 97 ... I34 UInsm n theV 6Mt 7931 845 843 6N 6N 6 ... 207 Cle m an th c 23243 2S257 27663 33817 36953 42328 ... S95 Wacu hplar 2a070 21 23? 280 307 353m ... 46 On pAic mntpns 1554S 1810 15661 23786 24t07 2382 .. 260 O privatt suew 4316 3722 4129 5S 540 11616 13tZ8 io Standrd (private) 3173 355 3666 5017 658 725 ... 91 hc1ast id assets 280 436? 22956 24146 3803 ... 46438 Clam on baks 358 421 7924 725S 10079 ... 1698? r.xud,d stochs 224 231 237 332 370 S8 ... 2202 rastwv accounts 263 3579 4U1 47 160 16321 20143 ... 2824 austs bsets Labilities 67757 51939 58520 9500 10665 118936 ... 157271 frerpJa k l lita- 1 1 6 6 6 ... I hiaw ad vnaqn depsts N570 32U9 3843 45 4855 MU62 ... S0748 km Pur 1882 26160 30513 35215 39367 4808 ... 38327 bingo Standard 5821 6678 7530 6784 962? 10654 ... 12m TMw SW tel) wr dts am 35 4 543 s50 m3 ... 7035 bso Popa 291 38W0 357 5 R6 S238 5611 ... 6865 elm Statd d 1% 10i 10 168 165 16 ... 170 6owrment deposits 62o 800 we10 13238 1836? 16567 ... 37UI tredt fro the 6 11 55" S 140 140 140 140 .. 174 twot m ts 2117 2260 244 2770 2772 2767 ... 388 fid alkiltta 2325 4336 5543 25501 267a3 326il ... 71550 its of other -.. .. tvaccw 27603 3402 364 5848 6248 7663 ... 6511 tershesb 34 656 1249 3832 3501 37MI ... 20491 eepayblt 86 11 - 261 4173 4S ... 13263 Ite dosts I t 161 - - - .. - in' luuhettes (let) *- .. .. .. .... imbilitUs 4? 66 330 1643 160 20752 ... 3125S Smute: hta roadd bp the bk of How6m. 1/ I, lako ff*q. -155 - ANNEX 2 Palte I of I ZN3JU$TL¢=:LI-E STUDY ftdlaq: SavH Ikas of tU d eua*q, 1913 (In ilhiw if uettuus) ........ .......................... . ............................ ..._.... ....._....... ., I8 1985 1986 1798 1989 breb Joe Sept. kc. Ilvch Jn - - -~~~~~. ............. _........ _. .............,,, rrp essets 3 2 3 61009 7 3O02 101 I S31 1294 Cl an n wm t 7207M 731 930 96 I1 1 3 11t"" 12M71 Clwens on te 73 9190m loom 1A 4 31 199law i60S i 137 Cl1a on tk by*s 1766 1500 873 7 3 7 2679 26 Whessafed ssets 77t26 351 719 17t6703 20173 2142011 ... 92 haded dstock 213 211 21 273 237 40 ... 92 Itdusti wto I3 - - 1271 1460 1439 ... 55s ruitorv yects 7681 7411 747 630m 719375 733M ... 100737 ONhrW sts -2 - 214 1133315 133 l0 1409 ... 191I93 lb tr ladilt.es 233963 2199 213222 53179 23702 2"125 337 1urn e mm 1 63 73 6167 133037 I1196 I5755P ... 2M2K6 Cwre c samed 2371 31137 33 31610 3 0 6 '7U 615" ber,' d4esits 6064 1693 16101 2959 3127 3m ... 41129 Dw deponts 2804 2"988 3626 263 1439 OM ... 96340 lift S freip cormiV4m ts ... ... 37 70 13103 130 ... 131 flebw and S 1 tern ferep liklilt V27227 248 36791 413130 143 U4341S 9191 612097 hart-tern faruap liklsihu 3030 26 3139 41 1716 S7117 1IS1 6Z350 o fkch - - - 609 11 192 23117 24410 6ousn t depsits W9910 477 20976 A1 414 90112 SO37 6260 tow counbts 9472 1 1 2 2X77 2000 3074 260 Wl fht lIabiltes 1831 76330 M73 1391190 169 1746197 ... 2393989 lestncted deots 2I00 21 1 20 00 2m00 2000 200D 2000 dit acwt - 1011 14 12914 1 9244 ...9400 fratlitarp ace ts 7S1% 7141 7179 22M3 269 2687 ... 3476 (ahe's dch 1635 IM 2664 7 I2 OS ... 1333 CatinWt lakiltuea (aft) - - . . _ . NW lidleities - 579 114091 1770 1457524 ... 19021 Wu; e0d Ptoaded b4 th h of IIonb,. _ 156 - ANNEX 3 Page I of 3 MOWZUBIQUE INDUSSRIAL SECTOR STUDY Structure of Interest Rates, September 1989 Ministerial Order Pursuant to the provision of numeral 5 of Resolution 11180 of December 31, concerning credit and interest rate policy, the Minister of Finance and the Governor of the Bank of Mozambique order the following: 1. The following interest rates are hereby fixed for lending transactions: I. Central Bank discount rate: 12 percent p.a. Percentage ceilings on use (percentage of demand deposits): (a) Discount: Ceiling up to 6 percent Term: Up to 3 weeks (b) Discount: Addition up to 6 percent Term: Up to 2 weeks. (In this case the rate shall be 2 points higher). II. Level 1 Covers the following economic sectors: - Agriculture and forestry - Farm marketing - Cooperative (agricultural, consumer and others) - Food industry - Electricity, gas, steam and water Working Capital Investment To To To To To Over 90 180 365 2 3 3 days days days years years years (in percent) 15 16 17 22 (single rate) _ 157 - ANNEX 3 Page 2 of 3 III. Level 2 Covers the following economic sectors: - Manufacturing (light and heavy) - Extractive industries - Construction, public works and housing - Export - Fishing - Railroad and maritime transportation and other public transportation of passengers and freight (enterprises and associations). Working Capital Investment To To To To To Over 90 180 365 2 3 3 days days days years years years (in percent) 20 22 24 25 26 27 IV. Level 3 Covers the following economic sectors. - Wholesale and retail trade, restaurants, hotels, tourism and other - Passenger and freight transportation (individual operations) - Other activities not included in Levels 1 or 2 and individuals. Working Capital Investment To To To To To Over 90 180 365 2 3 3 days days days years years years (in percent) 28 30 32 33 34 35 - 158 - AM=EX 3 Page 3of 3 2. Inivestments in construction and freight transportation made by [sic) the commerce and tourism sector and the housing sector by individuals may benefit from Level 2 interest rates. 3. Surplus inventories of essential goods in the consumer goods comzm'erce and industry sector may benefit temporarily from credit support at the rates set in Level 2. The Bank of Mozambique will determine by public announcement which goods are to be financed in these terms and the loan terms therefor. 4. The interest rate system fixed herein may also be applied to loans outstanding on the date of entry into force of this ministerial order if the pertinent contracts provide for modification of the interest rate in the event a different ceiling is legal established. 5. Disputes arising from interpretation and implementation ;f this [ministerial order shall be settled by] joint order of the Minister of Finance and the Governor of the Bank of Mozambique. 6. This ministerial order sha'l enter into force on September 6, 1989. Maputo September 6, 1989 Signed Signed Eneas da Conceicao Comiche Abdul Magid Osman Governor, Bank of Mozambique Minister of Finance - 159 - ANNEX 4 Page 1 of 1 HOZWMIQUE INDUSTRIAL SECTOR STUDY Distribution of bank credit to the economy. Mt billion. Sector 1984 1985 1986 1987 1988 1989 Total loans outstanding a/ (excl lending to Govt) 92.9 110.9 125.4 151.8 213.7 316.6 BM 72.0 87.4 99.2 119.6 156.0 213.9 BPD 18.8 20.7 22.4 27.3 41.4 62.4 BST 2.1 2.8 3.8 4.9 9.3 12.4 CCADR - - - 6.9 28.0 Agriculture 29.4 34.8 63.1 75.6 90.6 131.2 - BM 12.0 15.4 42.3 49.7 58.6 66.0 - BPD 17.1 19.0 20.7 25.2 26.7 38.4 - BST 0.3 0.4 0.1 0.7 0.7 11.5 - CCADR - - - - 4.5 25.7 Industry 38.1 47.1 21.4 23.9 45.6 75.0 - BM 35.1 43.9 18.1 20.3 27.2 51.4 - BPD 1.7 1.7 1.7 1.9 11.6 19.0 - BST 1.3 1.5 1.6 1.7 4.4 2.7 - CCADR - - - - 2.3 1.9 Construction 1.5 3.3 8.6 10.2 11.2 8.4 Transport & Communication 9.4 11.6 15.6 21.1 26.3 37.5 Domestic trade 6.6 6.6 16.9 11.0 26.5 48.6 Foreign trade 7.9 7.5 - 9.0 10.6 11.3 Other - 0.9 2.9 4.6 a/ Includes both domestic credit and foreign funds onlent by the banking system. Source: DIXF Bank of Mozambique. _ 160 - ANNEX S Page 1 of 1 MOZAMBIQUE INDUSTRIAL SECTOR STUDY Legal Status of Enterprises State Private Cooperative Intervened Mixed Total Petroleum 1 1 2 Paints 3 3 6 Toiletries 15 1 6 22 Industrial chemicals 8 2 10 Plastics 1 6 7 14 Rubber 2 3 6 11 Metal processing 1 6 7 14 Engineering 3 3 Ship repairs 2 1 3 Metal fabrication 2 5 1 9 17 Light engineering 2 45 4 8 3 62 Electrical 4 3 1 8 engineering Refrigeration 4 5 9 Construction 14 1 2 1 18 materials Cement 3 3 Glass 1 1 Wood products 8 10 1 19 Furniture 2 33 2 10 47 Paper cartons 1 2 4 Printing 3 15 13 31 Textiles 3 15 6 24 Garments 7 23 11 8 49 Leather 7 1 13 21 Sugar 2 4 6 Edible oils 5 2 7 Flour, 6 27 8 41 confectionery Salt 4 19 1 24 Cashew nuts 9 5 14 Meat products 3 3 Dairy 3 1 2 6 Processing 15 12 1 28 Drinks 8 14 22 Animal feeds 6 6 Other foods 3 4 4 11 Tobacco 6 6 Other 1 1 2 Pharmaceuticals 1 1 Total 114 294 21 140 6 5;5 Percentage of Total 19.8 51.1 3.7 24.3 1.1 100.0 Sources VNIDO - 161 - ANNEX 6 Page of Il MOZAMIQUE INDUSTRIAL SECTOR STUDY Main Problems Affecting the Firms Priority of Importance Problem 1* 2* 3* Code Definition No of S No. of X No of 2 firms firms firms pl Lack of raw materials (imported and local) and subsidiary inputs 30 55.5 6 11.1 1 19 p2 Lack of spare parts and tools 1 1.9 9 16.7 5 92 p4 Power cuts a/ 5 9.2 17 31.5 7 180 p5 Machinery (out of order, obsolete, old) 4 7.4 11 20.3 5 92 p6 Work force (lack of super- visory staff and skilled workers) 3 5.6 6 11.1 10 185 7 p Lack of transport equipment (for pasengers and goods) - - - - 6 111 p8 Technical assistance needs - - - - 5 92 p 9 Other 7 13 1 1.9 3 56 PIO No response 4 7.4 4 7.4 12 223 TOTAL 47 87.0 54 100.0 54 1050 Sourcet Responses of finns surveyed. UNIDO *Industrial Situation in Moambique. 1987 a/ This was a temporary problew of the economy with little impact at present, except In Beira. The conclusion to be drawn from the results obtained is that the management of the industrial firms surveyed considered the main problems affecting their operations to be the following, in decreasing order of importance. 1. Lack of raw (imported and local) subsidiary materials; 2. Power cuts; 3. Defective, old, and obsolete machinery with frequent breakdowns. - 162 - ANNEX 7 Page 1of 4 MOZSBIQUE INDUSTRIAL SECTOR STUDY IPORT TARIFF RATES - 1989 Product Rate X Minimum Maximum Animals on the hoof 4 25 Meat and parts, edible Free 34 Fish, crustaceans, mollusks 6 22 Milk, dairy products, eggs from fowls natural honey 10 24 Products of animal origin, unspecified 12 24 Live plants and floricultura products Free 34 Horticultural products, plants, roots and tubers, food Free 20 Fruits, citrus and melon rinds 10 22 Coffee. tee, mate and spices 2 27 Cereals 1 41 Milled products: malt, starch and potato starches, gluten, insulin 1 12 Oil seeds and fruits, miscellaneous seeds and fruits. industrial medicinal plants, straw and hay Free 25 Vegetable raw materials for dyeing and tanning gums, resins and other vegetable juices, and extracts 12 50 Materials for weaving and shaping and nonspecified products of vegetable origin 12 25 Fats and fatty oils, animal and vegetable: products from their breakdown, prepared food fats, animal or vegetable origin waxes 8 25 Meat, fish, crustacean and mollusk prepared foods 25 36 Nonspecified sweets and sugars 1 4 Cocoa and its prepared products 8 40 Prepared products from cereals, wheats or potato starches, pasta products 1 8 Prepared foods from horticultural products, fruits, and plants or plant parts 12 36 Miscellaneous food prepared products 12 Mt 250/kg Beverages, alcoholic liquids and vinegars 1 Mt 3000/lt Waste and discards from food industries, foods prepared for animals 10 24 Tobacco 6 30 Salt, sulfur, soils and stones, plasters limes and cements 1 35 - 163 - ANNX 7 Page I ofI4 Metallurgical ores, slag and ashes 10 10 Mineral fuels, mineral oils and products of their distilliation, bituminous materials, mineral waxes 5 10 Inorganic chemical products, inorganic compounds or organic compounds from precious metals radioactive elements, rare earth metals and isotopes 2.5 21.25 Organic products 8.75 20.85 Pharmaceutical products 12 25 Fertilizers 1.8 12 Tanning and dyeing extracts, tannin and its derivatives, coloring materials, paints and varnishes, mastic, writing inks 5 46.5 Essential oils and resinoids, perfume or toiletry products, cosmetics 10 50 Soaps, active organic products, preparations for lyes, lubricant preparations, artifical waxes, prepared waxes, products for maintenance and cleaning, candles for lighting and similar artifacts, modeling clays and wax for dentists 9 24 Abuminoid materials and glues 9 21.5 Powders and explosives, pyrotechnic articles, matches, pyrophoric alloys, inflammable materials 7 23 Photography and cinematography products 17 35 Miscellaneous products of the chemical industries 2.5 15 Artifical plastic materials, ethers and cellulose esters, artifical resins and products of such materials 5 25 Natural synthetic or artificial rubber and rubber products 5 25 Hide and skins 1S 23 Leather products, strap saddle and travel articles, purses, wallets, coin purses, cartridges and similar artifacts, gut products 11 40 Hides with fur for decoration and respective products, artificial hides with fur for decoration 5 50 Wood, charcoal and wood products 5 50 Cork and cork products 26.5 Mt 123/kg Mat and basket products 4.5 19 Raw materials for manufacture of paper 1 1 Paper, paper board and cardboard, cellulose pulp, paper, paper boa d and cardboard products 16 30 Bookstore articles and graphic arts products 1 23 Silk, silk refuse and silk waste 23 23 Synthetic or artifical textiles, continuous 23 23 Threads and yarns, with metals 40 60 Wools, furs and hairs 10 20 -164- ANNEX 7 Page 3 of 4 Flax and ramie 5 25 Cotton 5 25 Synthetic or artifical textiles, discontinuous 20 36 Other vegetable textile fibers, paper threads and respective tissues 5 20 Tapestry carpets, velvets, plushes, velvetized fabrics with loops and enflocked fabrics, ribbons, lace work, tulles, fixed mesh fabrics (netting), open work and seizing, embroicery 15 50 Pulps and felts, roping and other rope articles, special fabrics, impregnated or lined fabrics, technical textile material articles 5 35 Elastic mesh and respective artifacts 10 50 Clothing and clothing accessories, fabric 10 50 Other fabric artifacts 5 40 Used clothing, rags and cuttings 5 40 Footwear, leggings and analogous artifacts, parts of such objects 15 30 Hat wear and artifacts for similar use and respective parts 15 30 Umbrellas, parasols, canes, whips, long thin ,whips and their parts 10 35 Prepared decorative feathers and respective products, artificial flowers, fur products, fans 23 63 Stone, plaster, cement, amianthus, mica products and similar materials 5 36 Ceramic products 13 26.5 Glass and its products 10 30 Natural pearls, gems and similar items, precious materals, metals veneered with previous metal and respective products, false and ornamental jewelry 10 27 Coins Free Free Cast iron, malleable iron and steel 8 25 Copper 8 30 Nickel 8 15 Aluminum 8 15 Magnesium and beryllium (glucinium) 8 25 Lead 8 25 Zinc 8 15 Tin 8 15 Other common metals 8 25 Tools, flatware and cutlery, of common metals 8 35 Miscellaneous common metal products 8 35 Boilers, machinery, apparatuses and mechanical instruments 1 50 Machinery and electrical apparatuses and objects for electrical engineering uses 1 43 - 165 - ANNEX 7 Page 4 of 4 Vehicles and meterials for railways, nonelectric signaling devices for communication routes 1 2 Automobiles, tractors, velocipedes and other land vehicles Free 28 Air Navigation 1 1 Ocean and River navigation 1 24 Instruments and apparatuses and optical photographic and cinematographic products, measuring, verification and precision apparatuses, medical and surgical instruments and apparatuses Free 33 Clockworks 5 13 Musical instruments, apparatuses for recording and producing sound, parts and accessories of such instruments and apparatuses 5 30 Weapons and munitions Free 30 Furniture, medical and surgical furnishings, stuffed furniture articles and similar items 30 30 Materials for shaping or modeling, prepared or as a product 6 30 Brooms, brushes, mops, feather dusters, tassels, sieves and strainers 11 23 Toys, games and articles for recreation and sport 20 55 Miscellaneous products Free 50 art and collectible objects, antiques Free 18 Source: Government of Mozambique - Decree No. 20188. December 1988 3 - 1U6 _ ANNEX S INZAIrUEm DUS1UL SEC1IR SMWY EU'ORT R T RATS 1968 1907 1980 .I ishm.nt Pr.e.nt., .of Retention und Inuto oe 7cabiqiSe (state) 45 46 92 .,,.nhia 4do Cajo Monapo (private) 46 46 82 indo CXej (private) 46 46 82 -sheri.s) ocambique Pesca-tmopsca at.) 78 74 67 ,privte with mixed capital) 78 74 67 r (privat, with mixed capital) 78 74 67 (Private) -- 23 26 (private) 38 Rsia. Reis (private) 60 sO 60 coos Pese Cabrit. (privats) 46 -,.-ca (private with ixed capital) -- so _ca* do Sul (private) -- 8o 28 S. Cotton and sisal Cotton companies (tate) of Nampuls, Cobo Delgado, Inhambane, Softla, Mantco, Zamboxia, Tote and Niassa Ponding Pending 60 Coanhio Jose Forreirs do* Santos (private) 20 24.8 84 4. ME& and coen oiI Friiin d uiir (privtte) 20 20 46 Coprs Producers Association (various private and state nterpriet"s of Zambezi*) 20 50.7 87 Stat* Soereriat. for Light Industry end Food (govnmeant agency which subequently allocates tho resoures to the oil and soap Industries) s0 25 26 S. Tea i;che (state) 70 70 6 6. Other branches O of ctivity Usbor do Vocamoique (private with sixed capital) 100 100 tOO Lojis Frahass de llocmbique - Interfrrnc10 (state) - foreign currency *hops 100 100 100 Meditoc (stat.) - Pharmaceuticals, te. 100 100 100 magma (state) - seiprecious stone 100 100 100 II. fund of Soscific Pro cets (percentage of rec0Ipt 1n con vertab l currneie) Coto (state), Construcoes Teenicas (pri vato) and Conotrutora do Timego (private) - construction 100 100 100 Inflo s (state) - foretry 100 100 100 Samofor (mixed capital) - wood 100 100 100 Othor specific proJoctn (finane d by armarked grants frem MONAm , NAD, tho Netherlonds, SIDA, USAIO) 100 100 100 tII. Othr Funds (reulting from contractual agroment.) Companhl do Pipeline Mocasbique Zibabwb - CMPZ (privato with mixed capital) - petroloum products pipeline 100 100 100 Empress Necional de i4droatorbonets - ENN (stote) - petroleum 1OO 100 t00 Adminietranco da Posrquo Iobiliorto do Estado - APIE - housing 0o 80 30 Loans In foreign currency for lnvetment in exporting firm (experimental) 100 100 100 W0ue DalS" of Mo-s__ique I/Plus 10 perzent for permitted sals in freoly convertible currency in the local market. "02ASlQUE - DflJSTRIAL SECTOR SDY - FIXED CONSiER PRICES 1980 -APRIL71988 (In metical per kilogram or other unit. indicoted) ~~~ ~~~~ ~F*b-Mar 1l EJul-Aus 2/ J in Apr! 1980 1983 1984 1986 5sse 1987 19m Rofi ne Sugar 18.00 18.00 18.00 18.00 18.00 160.00 160.00 275.00 275.00 Unrefined Sugar 16.50 16.60 16.50 18.60 36.50 38.00 a3.00 60.00 264.00 Prawns (large) 100.00 38000 380.00 30.00 30.00 ... ... .../ ... y Rice, broken 4/ 7.00 21.00/7.00 21.00/7.00 35.00/14.00 62.60/21.00 62.60/21.00 138.00 Rice 'extra' 4/ 18.50 18.60 13.60 40.00/13.50 40.00/13.50 70.00/27.00 105.00/40.00 105.00/40.00 271.00 Maize 4/ 7.00 9.00 9.00 17.50/9.00 17.50/9.00 30.00/18.00 45.00/27.00 46.00/27.00 112.60 Mainl 7lour 4/ 9.60 9.50 9.50 22.60/12.50 22.50/12.60 40.00/25.00 60.00/37.50 60.00/37.60 145.00 Brad 7.00 7.00 7.00 7.00 7.00 66.00/45.00 55.00/46.00 60.00/90.00 116.00/162.00 Potetoe 9.00 9.00 9.00 60.00 ... ... Beana (butter) 16.60 27.60 27.60 34.50 84.60 l95.00 106.00 19S.00 260.00 Beans (regional) 10.00 14.00 14.00 24.00 24.00 128.00 128.00 128.00 198.00 Onions 12.60 12.50 87.60 ... ... ... Oroundnota 20.00 28.60 26.60 30.60 80.so 195.00 196.00 196.00 260.00 Edible Ci7 (liter) ... 38.60 56.60 65.50 58.60 240.00 360.00 360.00 540.00 Salt (common) 4.00 4.00 7.00 7.00 !/ 7.00 27.00 27.00 ... tJ *-- Fresh Milk (liter) 9.00 14.00 14.00 14.00 14.00 56.00 56.00 ... ... Condensed Milk (small can) 12.00 12.00 28.00 28.00 28.00 185.00 186.00 ... ... Eggs (dozen) 45.00 45.00 45.00 105.00 106.00 465.00 ... ... ... Beef (grade 1) 150.00 150.00 150.00 350.00 850.00 1,400.00 1,400.00 1,400.00 ... Pork (grade 1) 120.00 120.00 250.00 250.00 260.00 1,565.00 1,506.00 1,566.00 1,65C.00 Chicken 75.00 75.00 75.00 170.00 170.00 780.00 760.00 780.00 Fish (grade 1) 3.00 76.00 76.00 129.00 / 129.00 450.00 946.00 946.00 045.00 Fish (grade 2) 23.00 50.00 50.00 63.00 FJ 68.00 200.000 836.00 835.00 885.00 Water (i3) 6.00 6.00 12.50 12.50 12.50 52.50 78.75 160.00 160.00 Beer (60 ml) 86.00 39.00 39.00 86.00 86.00 240.00 855.00 855.00 355.00 Lemon Soda (260 ml) ... ... ... 10.60 10.50 86.00 85.00 Cigarettes (pock) .26.00 40.00 40.00 80.00 80.00 240.00 360.00 850.00 850.00 Matches (box) 1.50 1.S0 1.60 2.00 2.00 8.00 8.00 - - Soap 17.00 17.00 17.00 17.00 17.00 97.00 146.00 146.00 429.00 Butane (13 kg) 190.00 190.00 190.00 190.00 190.00 876.00 1,016.00 1,016.00 1,015.00 Petrol, Super (It) 27.00 27.00 27.00 27.00 27.00 ... 241.00 241.00 241.00 Lamp oil ... ... ... 10.00 10.00 38.60 66.10 66.10 66.10 - Bus Fare (Maputo) ... ... ... ... 8.50 10.00 :6.00 ... ... 0 Electricity (kwh) General ... ... ... ... 4.60 7.60 18.30 38.50 88.50 Residential ... ... 3.80 10.00 15.00 12.90 12.90 Battery 10.00 10.00 10.00 10.00 I .60 80.00 80.00 ... .. Sources: Ministry of Finance; and Ministry of Commerce lt Changes following first devaluation, including a few at end-January through as late as May. !/ Changes following second devaluation. 3/ Freed in September 1987. 4/ Where two prices are shown, the lower applies co Maputo (1986) or Maputo and Beria (1986-87). 5/ Prices ehown since May 1985 are for Maputo province. / Freed during 1937. - 168 - UNNEX t0 Page 1 of 1 MOUBIQW IRDtSThIAL SECTOR STDY Production Data Ministry of Industry and Energy (heavy LudustEy) (constant 1988 prices) million meticals 1987 1988 1989 Tyres 4102 2978 5528 Mechanical engineering 3490 5132 7896 of which: agriculture tools 355 549 1171 bicycles 777 1074 1110 Electrical Engineering 1288 1401 2039 Refrigeration 1798 2533 3311 Heavy Engineering 1256 2007 2149 Metal processing/products 3809 3419 5515 Paper products 327 313 396 Petroleum products 1434 1893 1578 Chemical products and allied 4956 5916 7928 of which: paints 2119 1804 2156 cardboard cartons 1528 1471 3249 Glass products 1402 2197 1200 Electricity Generation 7180 6999 6358 Total heavy industry & energy 31042 34789 43898 Sourcet Ministry of Industry and ENaergy. Department of Planning Maputo, November 1988 Notet Data as in original document. - 169 - ANNEX 11 Page 1 of 1 MOZAMBIQUE INDUSTRIAL SECTOR STUDY Scale Distribution of Industry - 1973 and 1987 Size category. Number of vorkers per enterprise. Total 1- 9 10-49 50-99 100-199 200-499 500-999 1000- 1973 Number of enterprises 1418 583 480 133 134 56 20 18 Total employment by size category (000) 99.5 2.1 12.4 9.5 18.4 18.0 13.0 26.5 1OOZ 1.8? 13.0Z 10.02 18.0 18.02 13.0? 26.5Z Total 1- 9 10-49 50-99 100-199 200-499 500 - 1987 Number of enterprises 2020 1145 551 134 98 64 28 Total employment by size category (000). 57.5 6.3 5.2 6.0 9.1 14.1 16.8 100l 11.02 9.02 10.6Z 15.82 24.62 29.2Z Scale Distribution of Industry: Haputo province - 1973 and 1987. Enterprise size by employment category. No employed Total 1-9 10-49 50-100 100-199 200-499 500+ 1973 No of enterprises 498 97 226 68 61 36 10 Approx employment 43350 450 6100 4700 8100 10800 13200 2 of employment 100 1.2 14.1 10.8 18.7 24.9 30.4 1987 No of enterprises 823 445 229 58 48 30 13 Approx employment 46400 2400 5430 4690 7560 11230 15080 2 of employment 100 5.2 11.7 10.1 16.3 24.2 32.5 Sources: based on Estatisticas Industriais 1973 and Ministry of Labor 1987 (unpublished, and excludes Zambezia province).