Report No. 32615-MZ Mozambique Country Economic Memorandum Sustaining growth and reducing poverty September 27, 2005 Poverty Reduction and Economic Management I Africa Region Document of the World Bank Contents EXECUTIVE SUMMARY ...................................................................................................... XI Introduction .................................................................................................................. xi The growth andpoverty record ................................................................................... xii Challenges to sustained growth andpoverty reduction ...................................... Challenges to growth andpoverty reduction: macroeconomic management ...........xvii Challenges to growth andpoverty reduction: private sector development ................xix Challenges to growth andpoverty reduction: natural resources............................... xxi .xxvi Summary of recommendations................................................................... Growth andpoverty outlook ..................................................................................... CHAPTER INTRODUCTION 1. ................................................................................................xxvii1 The context: a dramaticpost-conflict recovery..... ..................................... 1 A fast-growingpopulation .................................... 1 An overwhelmingly rural country .................................................. 2 A poor country, but lower poverty and inequality than the Africa average ..................3 Objectives and structure of the County Economic Memorandum................................ 4 CHAPTER 2. THEGROWTHRECORD ............................................................................... 7 A. Introduction............................................................................................................. .7 B. Growth and macroeconomic factors ........................................................................ 7 Tradepolicy ... ...................... 9 Resource mo ................................ 10 Publicjkance management..... .......................................................... 12 14 D. Examininggrowth from the supply side............................................................... C. Examining growth from the demand side.............................................................. .15 Growth in agriculture.................................................................................................. 17 Growth in manufacturing ............................................................................................ 19 Growth in infrastructure .............................................................................................. 20 .21 F. Idiosyncratic sources of growth............................................................................. E. A "growth accounting" approach.......................................................................... 22 Investment rates,foreign investment and skills ................................ Post-war catch-up and recovery of traditional markets Mega-projects ......................................... Foreign aid Combining the sources of growth ..... G. Conclusions.,, CHAPTER 3. POVERTY...................................................................................................29 A. Introduction and summary.................................................................................... -29 B. Trends inpoverty measures................................................................................... 29 Non-income weIfare trends.......................................................................................... 32 C. Accounting for the change inpoverty.............................................. Inequality.,............................................................................................ Determinants of income andpoverty, 2002/3........................... .. 37 Structure of income, livelihood strategies, and the labor ma .. 38 D. Conclusions............................................................................................................ 44 iii COUNTRY ECONOMIC MEMORANDUM CHAPTER PRIVATESECTORDEVELOPMENT: PRIORITIZATION 4. AND SEQUEKCIKG .47 A. 47 B. The private sector: adescription............................................................................ Introduction............................................................................................................ 47 50 D. C. Criteria for establishingthe prioritization................... ....................................... Improving access to finance and reducing its cost ................................................. 51 E. Improving access to and quality of energy............................................................ 54 F. Facilitating trade .................................................................................................... 56 ......................................... 56 .................................... 59 .................................... 61 G. .................................. 62 H. .................................. 63 Reducing the time and costfor registering a business ................................................ 63 Approving the new Commercial Code ......................................................... 64 Accelerating value added tax refunds ......................................................... 65 Reducing the cost of ingpections andJines while improving compliance.................... 66 CHAPTER 5 . THEROLEOFNATURAL RESOURCES INFASTER GROWTH AND POVERTY REDUCTION 69 A. Natural resources: characteristics incommon ....................................................... B. Ruralland............................................................................................................... 69 72 .................................................................................................... 75 .................................................................................................... 86 .................................................................................................... 90 F. Mining.................................................................................................................... 95 G. Conclusions.......................................................................................................... 101 CHAPTER 6 . GROWTH AND POVERTYOUTLOOK, AND KEY RECOMMENDATIONS .....103 A. Growth inthe long tenn....................................................................................... 103 (a) A regional comparison and a general equilibrium mod ....................... 103 (b) An intercountry comparison using the "growth regves erature ..........106 (e) Future growth and institutions.. ....................... 107 B. A briefexaminationo fthe growthprospects inkey sectors................................ 110 Agriculture.......................................... ........................ 110 Manufacturing .................................... ....................................... 114 . Tourism................................ ....................................... 116 Prospects for future poverty reduction ................................................................. 117 D. Keyrecommendations ......................................................................................... C 118 REFERENCES ................................................................................................................... 121 ANNEX1 TABLES FIGURES AND ................................................................................. 135 ANNEX2 BACKGROUND PAPERS FOR THE CEM ....................................................... 162 Figures Figure 1. The real effective exchange rate (1980=100. .depr.)................................................... 9 Figure 2. Foodcrop output, 199617-200314............................................................................... 18 iv Figure 3. Selected sub-Saharan countries: Net aid as a percentage o f GDP. and annual per capita aid. average 1983-2002............................................................................................... 25 Figure 4. Food crop production and rural poverty. 1996-2003 .................................................. Figure 5. Real interest rates paidby manufacturing firms - intercountry comparison..............37 52 Figure 6. Natural disasters and GDP growth, 1980-2004 ........................................................ 105 Figure 7. Maize yields, Mozambique and its neighbors, 1961-2003....................................... 111 Tables Table 1. Demographic projections. 1990-2050............................................................................ 2 Table 2. Economic and social indicators for Mozambique and sub-Saharan Africa ........... Table 3 Basic macroeconomic indicators .................................................................. Table 4 Macro framework, 2003-2006 ............................................................................... Table 5... Growth o f GDP components, 1992-2003 .............................................................. Table 6. Sectoral breakdown o f GDP, 1996-2003 ......................... ....................................... 15 Table 7. Gross domestic product by sector: real growth rates (percent) .................................... 16 Table 8. Real industrial production index by subsector (1999=100) ......................................... 20 Table 9. Sources o f economic growth from a Solow-Denison model (percent) ........................ 21 Table 10. Table 11. Poverty headcount by province (percent) .................................................................. Indicators: health, education, water, 1990-2003 ........................................................ 26 Table 12. Percentage change inpoverty rates, adjusting for urban/rural border changes .........30 31 Table 13 . 32 Non-monetary measures of welfare (percentage) ...................................................... Growth o f consumption and pro-poor growth ........................................................... Table 14. 32 Table 15. Access to services ...................................................................................................... Table 16 Selected health outcome indicators.,..................................................... ......33 34 Table 17 Growth o f consumption by quintile, 199617-200213 (percentage) ............................. 34 Table 18 Measures o f inequality. 199617 and 200213 .............................. ............................ Table 19 Table 20 ..... 36 Household compositionby consumption quintiles .................................................... Decomposingpoverty changes (geographical and sectoral dimensions) ...................35 Table 21. Employment by sector and by ruralhrbanlocation, 199617 and 200213 ...................38 39 Table 22. Average annual growth rates, employment and value added, by sector, 199617to 200213.................................................................................................................................... 42 Table 23. Employment status by consumption quintile and location. 199617 and 200213 ........ 49 Table 24. Firms'perceptions of "major" or "severe" obstacles to business. 2001 .................... 50 Table 25. Exports by main category .......................................................................................... Table 26. Exports by country o f destination, 1998-2002 (inpercent o f total exports) ..............57 58 Table 27. Imports by country o f origin, 1998-2002 (inpercent o f imports) .............................. Table 28. Summary o f business climate constraints and proposed recommendations ..............59 68 Table 29. Current and potential rents from naturalresource use ............................................... 71 Table 30 Expenditures in selected sectors, 2002-2004 ............................................................. 77 Table 31 .. Water indicators for Maputo city, 2000 and 2004 ..................................................... Table 32. Quotas and catches - shallow water shrimp trawlers with freezers, 1995-2003 .......82 87 Table 33 National wood production (m') 1998-2003 ............................................................... 90 Table 34 91 License fee revenues, 1998-2003 ............................................................................... License fee structure per cubic meter, 1998-2005 (US$) .......................................... Table 35 91 Table 36 .... Value o f annual mineral production (US$) ................................................................ 95 Table 37. 96 Natural resources: summary o f findings and recommendations.............................. Poverty rates by sector o f employment o f the householdhead ................................. Table 38. 101 Table 39. Poverty rates, projections for agriculture and non-agriculture ................................ GDP growth rates o f Mozambique's neighbors ...................................................... 104 Table 40. 118 V COUNTRY ECONOMIC MEMORANDUM Annex Tables and Figures Annex Table 1 Selected economic and financial indicators .......... ................................ Annex Table 2 .. Mozambique's progress towards the Millennium opment Goals ........135 Annex Table 3. 137 Annex Table 4. Employment by employer type, 1996/7 and 200213 (% o f labor force) .........136 137 Annex Table 5. GDP per capita, and GDP shares by provincelregion, 2000 .......................... Growth o f GDP by provincelregion, 1997-2000 ............................................ 138 Annex Figure 6. GDP per capita in 1996 and annual growth o f GDP per capita 1996-2000, by province ................................................................................................................... Annex Figure 7. GDP per capita in 1996 and annual approved investment 1996-2000 as a percentage o f GDP, by province ................ ................. Annex Table 8. Macroeconomic impact o f the mega-projects, 1998-2002 Annex Table 9. Projected macroeconomic impact o f the mega-projects (2003-2010) ............140 Annex Table 10. Distribution o f projectized foreign aid by sector ...............140 Annex Table 11. Non-monetary measures o f welfare ....................... 141 Annex Table 12. Access to services (percentage) ................................................................... 142 Annex Table 13. Health outcome indicators................................... .......................... 143 Annex Figure 14 . .......................... 144 Inequality in sub-Saharan Africa ................................................................ Growth incidence curve, 199617to 200213........ Annex Figure 15. 145 Annex Table 16. Decomposing changes inpoverty inducedby changes inmean consumption 146 Annex Table 17. and inequality (by geographical and sectoral dimensions) ................................................. Decomposition o f the change inpoverty by geographical and sectoral dimensions .......................................................................................................................... Annex Table 18. Household head employment sector and contract (1996/7 and 200213) ...... 147 148 Annex Table 19 . Consumption regressions ............................................................................. 149 Annex Figure 20. Sources o f cash income (excluding private transfers) and self-consumption, by location, 200213.............................................................................................................. 150 Annex Figure 21. Sources o f cash income (excluding private transfers) and self-consumption, by quintile, 200213 .......... ............................................................................................... 151 Annex Figure 22. Poverty rates (based on per adult equivalent consumption) by sector of employment ofthe household head..................................................................................... Annex Table 23. Employment by sector, by ruraliurban location and by poorhon-poor .......152 Annex Table 24. Sector employment rates, by sex and urbanlrural (percentage) ...................153 153 Annex Table 25. Type of contract, by industry and by urbadrural in2002 (men and women) 154 Annex Table 26. Type o f contract, by industry and by urbadrural in2002 (men).................154 Annex Table 27 Type o f contract, by industry and by urbanhral in2002 (women) .............155 ............155 Annex Table 28 .. Type of contract, byindustry, location and poorlnon-poor, 200213 Annex Table 29. 156 Firing costs and firmperceptions insub-Saharan Africa ............................. Comparative labor regulations and worker benefits .................................... Annex Table 30 . Annex Table 31. Difficulty o f firing and firm perceptions insub-Saharan Africa Annex Table 32 . ..................157 157 Wage regressions, including agricultural workers, 200213 .......................... Annex Table 33. Commodity composition o f imports, 1997-2002 (US$ millions) ................158 159 Annex Figure 34. Access to bank loans by manufacturing firms, I C A countries.................... 160 Annex Figure 35. Claims on the private sector as a fraction o f GDP, international comparison, 2001..................................................................................................................................... Annex Figure 36. Value o f collateral required for loans, intemational comparison................160 161 Table of Boxes Box 1. Summary o f Growth and competitiveness ..................................................................... 51 vi Box 2.Summary ofRural development strategy..................................................................... 112 Box 3.Summary of Contractfarming and supply chainjnancing ......................................... 113 Box 4.SummaryofImpacts of extension sewices .................................................................. 114 vii COUNTRYECONOMICMEMORANDUM ABBREVIATIONS AND ACRONYMS AGOA Africa Growth and Opportunity IRPC Impost0 sobre o rendimento das Act pessoas colectivas (company tax) CACM Arbitration, Conciliation, and MADER Ministerio da Agricultura e Mediation Center Desenvolvimento Rural CEM Country Economic Memorandum MDGS MillenniumDevelopment Goals CGE Computable general equilibrium MPF MinistryofPlanning and Finance model NGO Non-Governmental Organization CPAR Country Procurement Assessment OECD Organization for Economic Report Cooperation and Development CPI Current Price Index PARPA Programa de A950 para a Reducg5o CRA Conselho de Regula@o do de Pobreza Absoluta Abastecimento de Agua PRSP Poverty Reduction Strategy Paper (Regulatory Board for Water PPP Purchasingpower parity sUPP1Y) QUIBB Questionario de IndicadoresBasicos DHS Demographic and Health Surveys de Bem-Estar (qualitative EBA Everything-But-Arms initiative indicators survey) EDM Electricidade de Moqambique PES Plano Economico e Social (Social EU EuropeanUnion and Economic Plan) FY Financial Year RBMMP Roads and Bridges Management and EPA Economic Partnership Agreement Maintenance Program, sector-wide (with the EU) program, sometimes referred to as FIPAG Asset Holding and Investment "Roads 111" Company for the five cities' ROADS I11 See RBMMP water ROCS Roads and Coastal Shippingprojects GDP Gross Domestic Product 1 and 2, with World Bankand GNI Gross National Income donor participation GOM Government o f Mozambique RPED Regional Program for Enterprise HIPC Highly IndebtedPoor Countries Dynamics (at the World Bank) initiative SISTAFE Sistema Integrado de AdministraGEio IAF Inqukrito aos Agregados Financeira do Estado (integrated Familiares financial management information ICA InvestmentClimate Assessment system) ICOR Incremental Capital Output Ratio SSA Sub-Saharan Africa IFPRI Intemational Food Policy TFP Total factor productivity Research Institute TFR Total fertility rate ILO Intemational Labour Office TIA Trabalho de Inquerito Agricola IMF Intemational Monetary Fund (agricultural surveys) INE Instituto Nacional de Estatistica UNDP United Nations Development INNOQ Instituto Nacional de Program Nomaliza@o e Qualidade VAT Value added tax WB World Bank .., Vlll ACKNOWLEDGEMENTS This i s the white cover report o fthe Country Economic Memorandum (CEM) undertaken by the World Bank and the Ministry o f Finance o f the Government o f Mozambique. Useful information and insight was provided by experts inmany departments inthe Ministry o f Finance. For the World Bank, the C E M was managedby Peter M o l l (AFTP1). The team included Asmara Achcar (WBIHD), Luiz Gabriel Azevedo (LCSES), Elena Bardasi (AFTPM), Maria Teresa Benito-Spinetto (AFTPl), Gilbert0 de Barros (AFTPS), Paul0 de Sa (LCCSC), Louise Fox (AFTPM), Antonio Franco (AFTPl), Jose Luis Macamo (AFTPR), SCrgio Margulis (AFTSD), and Katleenvan den Broeck (AFTPM). Contributions were made by the consultants Alvaro Carmo Vaz, Arne Eide, Gordon Hughes, Stephan Klasen, Ana Menezes, Jose NegrBo, Ronald0 Seroa da Motta, Almeida Sitoe, Joaquim Tenreiro de Almeida and Silke Woltermann. Part o f the funding for the consultants came from the Belgian Poverty ReductionPartnership Program Trust Fund and the Norwegian ESSD Trust Fund. Assistance with editing, communication, budget management and arranging o fmeetings was provided by Ligia Murphy (AFTP1) and Lurdes Malate (AFC02). Onthe Government side, the review was managedby Mr.Pedro Couto, former Director, and Mr.Adriano Ubisse, Deputy Director, o f the Gabinete de Estudos (Research Department) o f the Ministryo f Finance. Especial thanks go to the then Minister o f Planning and Finance, Ms.L.Diogo, and the President o f the Instituto Nacional de Estatistica, Dr.J. Loureiro, for their assistance. Many government ministries and institutions collaborated, including members o f the Banco de Moqambique, the Instituto Nacional de Estatistica, the Instituto Nacional de Meteorologia, the Ministry o f Agriculture, the Ministry o f Communication and Transport, the Ministry o f Coordination of the Environment (MICOA), the Ministryo f Education, the Ministry o f Fisheries, the Ministryo fLabor, the Ministryo fPublic Works, the Ministry o fTourism, and the Ministryo fTrade and Industry. Advice and comments were also received from the economists o f the G17 group of donors, the Association o f Mozambican Economists (AMECON), the organizers o f the Agenda 2025, the private sector as representedby CTA, several privately owned firms and associations o f firms, the study unit Forum Terra, and the G20 group o f NGOs (including Grupo da Divida) and trade unions. Their extensive knowledge o f economic issues inMozambique was extremely valuable. The peer reviewers, Zeljko Bogetic (AFTP l), Martin Ravallion (DECRG), Robert Schneider (LCSES), Zmarak Shalizi (DECRG), Dominique van de Walle (DECRG) and Nancy Vandycke (ECSPF), provided valuable comments and suggestions at different stages o f the report. The report was preparedunder the supervision o f Emmanuel Akpa, Sector Manager for AFTP1, H e offered overall conceptual guidance, providedcritical analytical advice and ensured quality control and management support. Michael Baxter, Country Director, supported the process and providedthe major guidelines. ix EXECUTNE SUMMARY EXECUTIVE SUMMARY Introduction 1. Mozambique has staged a dramatic recovery from the damage o f the civil war. Since 1992,infrastructure has been improved and i s now approaching its pre-war levels, and incomes have risen considerably. The poverty headcount fell from 69 percent in 1996/7to 54 percent in2002/3. On average the economy grew by 8 percent annually between 1996 and 2003. This accomplishment can be attributed to the Government's phased but determined approach to stabilization and structuralreforms, as well as to concessional assistance (half o f Government expenditures), a remarkable agricultural "catch-up", an expansion o f agricultural exports, and fast expansion intourism, construction, and certain manufacturing subsectors. Another factor was the authorities' success inattracting "mega-projects" inaluminium smelting, natural gas, and titanium mining,and a resultingtripling o fexports inthree years. ii. Another factor which was a precondition for all o f the above i s the fact that the country was successful inbringingabout reconciliation, ending the civil war, and in managing potential conflicts since that time. Mozambique has just had its third general and presidential election. ... 211. Nevertheless the country remains poor ($240 per capita income); infrastructure i s inadequate, there are serious unmet education and health needs, and poverty rates remain high. Many o f the "first-generation" reforms associated with market liberalization have already been implemented. The country now faces the prospect of tighteningmacroeconomic constraints, a need for substantial institutional improvement to make growth sustainable, a increasing need for better prioritization and management o f public expenditures, to eliminate absolute poverty, and massive investment in infrastructure to promote growth and further reduce poverty. iv. Objective and structure of the Country Economic Memorandum. This Memorandum has examined the growth-poverty linkage, using a wide variety o f data sources, includingthe recently completed national household survey (2002/3). It has sought to understand the sources o f growth inthe recent past, to evaluate the prospects for growth inthe next decade, to examine the likely implications for poverty, and to outline the policies that will be needed to achieve further growth and poverty reduction. The reason for this focus at this time i s that the Government was planning to develop its second Poverty Reduction Strategy Paper (termed PARPA inMozambique) and it was felt that a substantial contribution inthe area o f growth and poverty might make a useful contribution. v. Inaddition-on account o fthe light coverage the subject has received inother sector work, Bank-produced and otherwise - the Country Economic Memorandum also examines the relevance o f natural resource management to growth and poverty objectives. It was found, inthe event, that natural resources are o f great importance for both growth and poverty reduction. Hence it i s hoped that inthe future the natural resource issues will be fully integrated with the national policy discussions. x i COUNTRY ECONOMIC MEMORANDUM vi. After the preparationo f the C E M had begun, the Government made an additional request, namely that it include a chapter on private sector development. This had not formed part o f the original plan, because a great deal o f valuable research had already gone into this important subject (as i s documented inChapter 4, p. 47). Hence it was agreed that the contribution o f this Memorandum would not be to break new ground in the area o f private sector development, but to arrive at a workable synthesis and prioritization o f the many disparate recommendationsarising from these efforts. vii. This Executive Summary pulls the arguments together inthe following sections: P The growth andpoverty record k Challenges to growth andpovertyreduction:private sector development P Challenges to growth andpoverty reduction: natural resources P Growthand poverty outlook The growth and poverty record viii. Mozambique's population o f 19 million is projected to grow at about 2.4 percent annually between 2005 and 2010. Since dependency rates are falling, and the country i s one of the most land-abundant inAfrica, population dynamics have not been a major driver of poverty trends; indeed, demographic dynamics have helped support rising per capita incomes and falling poverty. The urban working population i s expected to grow at 4 percent until 2010, underscoringthe need for a growth path withjob creation. IX. How was 8percent growth accomplished? The preconditions for healthy growth were met with a tolerably stable macro environment, steady progress towards liberalizing key sectors (for example telecoms, air transport), and steady progress in freeing up intemal and external trade. Inaddition, there were several idiosyncratic factors includingmassive concessional assistance. These are discussed inthe ensuing paragraphs. X . As concems the macro environment, inflationwas brought down from over 60 percent to single digits inthe late 1990s, and despite some slippages inthe wake o f the banking crisis in2000-2002, i s expected to remain inthe single digit range inthe future. Government spending increased markedly (24 percent o f GDP in 1997 to 30 percent in 2002, then falling to 24 percent by 2004) but there was little borrowing from the banking system as the spending was hndedfrom increased concessionary assistance (andfrom a modest increase inrevenues). Since 2001, the Government's fiscal and monetary stance has been guided by the Program o f Action for the Reduction o f Absolute Poverty (PARPA), which stresses the key role o fmacro stability, envisages steady increases in revenue, and brooks no borrowing from the banking system. xi. As concems liberalization, state-sponsored interventions inthe agricultural economy were all but eliminated by the mid-1990s. Telecoms were liberalized significantly from 2001 onwards and cell-phones increased from 51,000 in2001 to 700,000 by 2004. Air traffic was liberalized in2002, and 2004 saw the first new entrant competing on domestic routes, and the first tourist charter flights. xii. As regards trade, through the 1990s, import tariffs were steadily lowered to the point where average tariff i s 9 percent, one o f the lowest inAfrica. The plan i s to reduce the top tariff rate to 20 percent in2006. Management o f customs was contracted out xii EXECUTIVESUMMARY during the 1990s, and the efficiency o fcollection improved. Export performance has been strong, growing at 22 percent annual growth (inUS$ terms), well above the performance o f world exports (6 percent), making Mozambique one of the few countries inAfrica whose shareinworld exports hasrisen. Most ofthe export expansionwas, however, due to mega-projects; traditional exports grew at only 2.3 percent. xiii. Inaddition, Mozambiquebenefitedfrom three idiosyncratic factors: > Agriculture, which supports 80 percent o f the population, grew at 6.8 percent per annum between 1992 and 1997, slowing to 4.6 percent per annum between 1997 and 2003 (including the impact of the floods o f 2000). The main contributors are foodgrains (maize), sugar and tobacco. About halfo f the growth between 1992 and 2003 was due to area expansion (2.4 percent annually), and the remainder labor force growth and yield improvements. Growth incertain regions such as Tete has been stronger owing to trade. A substantial part o f the growth i s due to a "bounce-back", intwo senses. Some o f it took place as crop yields, which had fallen precipitously duringthe socialist period and the war, quickly recoveredto their original levels during the 1990s. Inthe other sense, during the course o f the 1990s the areaharvested rose back to its original pre-war levels, as war rehgees returnedto their farms. Inaddition, off-farm income (micro-enterprises, wages and remittances) were important inthe three southernmost provinces which dependon links with Maputo city. > Mozambique i s one o f the bigrecipients o f concessional assistance inAfrica (some 12-15 percent of GDP, accounting for half of Government spending). The aid succeededin assisting the Government to pursue its aims o f improving access to and quality of services -to judge by the improvements ineducation, health and roads - and in improving somewhat the level o f fiduciary accountability. (An exception i s HIViAIDSwhich continues to spread.) Modeling approaches show that aid was responsible for a substantial part o f Mozambique's impressive growth record. k Another factor isthe mega-projectphenomenon. Attractedbyfiscal benefits and natural resources, several very large investors entered Mozambique after 1997, starting with the M O Z A L aluminium smelter. By 2002, these accounted for 7 percent o f GDP. Their contribution to GDP growth was perhaps as much as 1.6 percentage points annually. Their share o f GDP will rise to 10-11percent by the end o fthe decade. The megaprojects helped"put Mozambique on the map" and will probably encourage non-megaprojectinvestment also. All these effects have beenpositive to a greater or a lesser degree, and no negative effects havebeen observed. Hence there i s at least aprimafacie case that the authorities' unorthodox policy of tax incentives has been successful. xiv. All this meant that growthwas broad-based. Inadditionto agriculture, there was substantial growth insome parts o f the private formal sector -manufacturing, transport, communications, services, mining-where average growth rates (inreal terms) between 1995 and 2003 variedbetween 5 and 15 percent annually. Some o f this was due to the reestablishment o f traditional markets such as tourism from the region, hotels and restaurants. Some constituted new investment, much o f it fiom neighboring countries. This was growthfrom a very low base and so these rates are not likelyto be as highinthe future. The number ofpeople active inthe urbaninformal sector grew at the fast clip of xiii COUNTRYECONOMIC MEMORANDUM 7-8 percent, driven by rural-urban migration and fast growth o f the working age population. xv. These observations should be tempered by noting that this remarkable growth spurt occurred despite an unpropitious institutional environment inmany respects (see Chapter 4): a weak judicial system, an inability to enforce contracts or retrieve collateral, a troubled financial sector, undeveloped "civil society'' institutions, l o w civil service capacity, and poor (though improving) infrastructure. On the other hand, there were compensating institutional factors: a fairly unified government with a firm commitment to poverty reduction, an ability to embark o n reforms as needed, the absence o f armed conflict, progress inlowering trade barriers, and free entry into agricultural marketing. xvi. How didpovertyfall? Usingthe Government's poverty line, the poverty headcount fell dramatically, from 69 percent in 199617 to 54 percent in200213. The fall was faster inrural areas (from 71 to 55 percent) than inurban areas (62 to 52 percent). Nonetheless, poverty continues to be primarily a rural phenomenon because the bulk o f the population is rural. xvii. Poverty can also be measured by assets, access to services, and social outcomes. These also improved. Radio and bicycle ownership rose. Safe water access rose from 24 percent to 3 7 percent. As o f 199617, 8 percent of all adults had completed primary school (EP2); this rose to 11percent by 2002/3. The key health indicator is the infant mortality rate, which fell from 149 in 1995 to 101 in 2003, one o f the fastest reductions observed in Africa. xviii. More generally, consumption as measured by the household surveys increased at an annual average rate o f 4.6 percent. Furthermore, consumption rose throughout the income spectrum. For example, the consumption o f the poorest 20 percent o fthe population rose by 23 percent between 1996/7 and 2002/3, or 3.5 percent annually. xix. As for the regions, poverty decreasedinall provinces but Cab0 Delgado and Maputo province. These latter anomalous findings may, however, be discarded. That for Cab0 Delgado i s due to sampling errors in the household surveys. That for Maputo province is due to the fact that the food consumed inthe area i s more expensive, being o f higher quality.' Ifit were possible to standardize for food quality, the poverty rate in Maputo province might well havefallen between 199617 and 200213. xx. Several policy-relevant factors can be identified as having had considerable importance inreducing poverty between 199617 and 200213: 9 Fromthe sectoral point ofview, the dominant factor inexplaining Mozambique's poverty performance is the good performance o f agricultural households. This i s because o f the vast size o f the agricultural sector relative to other sectors o f employment. 9 The increase of educationhelpedto reduce poverty. Thisisbecause the levels of education rose, and because the impact o f education o n wages (and consumption) i s considerable. The impact o f an extra year o f education o n the wage is o f the order o f 3 percent inagriculture inrural areas, and between 5 and 15 percent in non-farm activities inrural areas, and also inurban areas; and the impact rises at the ES2 and tertiary levels. 'See the maintext (page 3 1, footnote 64) for further discussion of these data. xiv EXECUTIVE SUMMARY xxi. Linkages between growth andpoverty reduction. H o w did the growth process contribute to reducing poverty inthe period 1996 to 2003? First and most important i s to note that while GDP grew at 8 percent in the period 1992 to 2003, private consumption as measured by the national accounts grew at a more modest 2-4 percent, depending on the specific time periode2(The wedge between these was due to rising investment rates and steeply increased exports associated with the megaprojects.) It was this consumption growth that led to the observed reduction inpoverty. xxii. Second, it i s worth noting that growth was "pro-poor". For the purposes o f the Country Economic Memorandum, pro-poor growth i s defined as the mean rate o f growth o f consumption o f people below the poverty line. This "rate o f pro-poor growth" i s strongly positive (3.9 percent per year). No important region or occupation was left out o f the growth process. xxiii. The specific linkages between growth and poverty reduction can now be enumerated. As mentioned above, growth in agricultural output, partly due to area expansion and partly due to improved productivity, reducedpoverty directly inrural areas (and some urban areas). Non-farm pursuits inrural areas likewise helped to reduce poverty directly, particularly inthe three southernmost provinces where there were closer market linkages. The increase inconsum tion that i s observed at all levels i s financed by increasing income. Employment income rose considerably between 1996/7 and 2002/3 : P the average household's employment income was Mt 5,50O/day in 1996, and this rose to Mt 15,000 in200213.4 xxiv. As the economy grew, structuralchange occurredinthe expected ways. Value added grew faster inservices and consequently employment growth was faster in services, Value added grew more slowly inagriculture and consequently employment growth inagriculture was lower, and people moved from rural areas to urban. xxv. Duringthisperiod, Government spending-backed upby donor support - focused on the PARPA priorities o f education, health, roads, water and agriculture. Each of these spending areas helped to reduce poverty, directly or indirectly: 9 Increasededucationspendingraisedthe (gross) enrolment rateinlowerprimary schooling from 56 percent in 1995 to 110percent by 2003. The completion rate for EP1rose from under 20 percent in 1990 to 40 percent in2003. The number o f adults with completed primary school rose from 8 percent in 1996/7 to 11percent in2002/3. The superior educational levelspermittedbetter earnings inthe labor market and raised the productivity o f informal sector operators. 9 Healthspendingincreased accessto healthfacilities and assistance (e.g. increased vaccination coverage), and to this extent reduced poverty directly. Inaddition the improved health status o f people has improvedtheir productivity, which presumably helped to reduce poverty indirectly, as working adults are now more likely to be able to obtain medical care when it i s needed, and the total fertility * These figures are lower than the measure from the household surveys o f 199617 and 200213. This does not necessarily imply that either i s mistaken. See the discussion infootnote 45, p. 14. "Employment income" here refers to the sum o f formal and informal employment income. The precise shares of the two are unknown. See footnote 4. These figures are inconstant 2002 Mt. Problems o f definition beleaguer the measures o f informal income, and hence also total income. Hence here we stress: (a) consumption, and (b) employment income. xv COUNTRYECONOMIC MEMORANDUM rate has fallen inresponse to the falling infant mortality rate, thus freeing more women for market work. Efforts to stem the HIV/AIDS epidemic have not yet succeeded, however (see paragraph xxvii below). P With spending on road rehabilitation andmaintenance, between 1992 and 2003, roads ingood or fair condition rose from 10 percent to 70 percent o f the classified network, and impassable roads decreased from 50 percent to only 5 percent. The impact on poverty is direct, via employment inroad rehabilitation and maintenance efforts, and also indirect, via cheaper transport, lower margins, more trade, better farmgate prices for farmers, and lower prices o f consumer goods in remote areas. P Water spending raised access to safe water from 12 to 27 percent inrural areas, and from 54 to 64 percent inurban areas, between 1996/7 and 2002/3.5 This likely reduced the amount o f water-borne disease (e.g. diarrhea and cholera). P Agricultural extension, and also encouragement o f outgrower schemes (cotton, tobacco, oil seeds, vegetables) helped to drive agricultural GDP growth o f 5.6 percent annually between 1992 and 2003. xxvi . Intwo areas, the links between growth andpoverty reductionwere not as strong as might have been expected: P The formal sector created little employment. The growth elasticity o f formal sector employment was 0.1,6 low compared with other countries. A factor behind this may be the restrictivelabor legislationwhich makes it unattractiveto expand employment. (See further comment o n the labor law at para. 120ff, p. 40). P The mega-projects contributedgreatly to value added, buthave so far had a small (though growing) impact on employment or on local business via backward linkages. By 2010 the mega-projects will probably account for some 2 percent o f private sector employment, and it i s expected that backward linkages will increase inthe future. The main contribution o fthe mega-projects to poverty reduction is via their tax contributions, which are expected to rise to 3 or 4 percent o f Government revenues by 2010. This reinforces further the point that government expenditures needto be made as efficient as possible so as to make good use o f these increased revenues. xxvii. And finally there is one area o fpoverty-relevant outcomes inwhich neither growth nor Government efforts have been able (yet) to stem a deterioration: HIV/AIDS. Prevalence o f the disease has risen remorselessly, from 12 percent in 2002 (Amdt, 2003) to 16.2 percent in2004. The impact o f the epidemic is likely to grow more serious duringthe next decade. By 2010 life expectancy at birthis expected to drop from 43 to 36 years, rather than increasing to 50 years. The macroeconomic and social costs are large. A modeling approach indicates that the advance o f AIDS reduces per capita GDP growth by as much as 1percent per annum, because o f reduced productivity growth, 5Figures from the household surveys. See the discussion inparagraph 270 (p. 80) and footnote 128 (p. 80) . about measurement o f access to clean water. This refers to firms unaffected by restructuring and privatization o f state enterprises inthe mid-1990s. See the discussion inparagraph 120 (p. 40), and footnote 85, as well as and Benito-Spinetto and M o l l (2005) for details. xvi EXECUTNE SUMMARY reducedhumancapital accumulation, and reduced physical accumulation, not to mention the dramatic social burden o f the disease. xxviii. The Government i s seeking to respondto the challenge. There i s a national HIVIAIDS strategy inplace, as well as a national HIV/AIDSsecretariat. HIV/AIDS mitigation plans are being executed inthe major ministries. There i s no lack o f donor funding. But execution rates have been low-executionwas only 70 percent o fplanin 2004 and of the World Bank-funded HIVIAIDSproject only US$ 5 million was spent in the first two years. Better training and management i s urgently needed. Challengesto sustained growth andpoverty reduction xxix. Mozambique appears to be well placedto experience a long period o f sustained per capita growth, and also to achieve some o f the key Millennium Development Goals, such as the expected halving o f the poverty rate by 2015. Nevertheless there are many remaining challenges. There remains the possibility that with swifter progress incertain key reforms, the growth rate couldbe as high as 7 or 8 percent inthe long term, and that poverty reduction could be quicker. The likely growth scenarios for the future - and the correspondingpoverty scenarios - are discussed below (para. lxxx, page xxvi), but before speculating on the overall growth rate it i s essential to examine the factors that will influence growth patterns inthe future, and consider the Government's role inwarding off threats to continued growth, as well as its role inseizing new opportunities for growth. The following sections explore these. xxx. The challenges facing Mozambique's decision makers can be divided into three categories: macroeconomic management, private sector development, and natural resouyces. Challengesto growth and poverty reduction: macroeconomic management xxxi. Itwas mentioned above that Mozambique's tolerably good macroeconomic performance was one o f the pillars o f its excellent growth record inthe past decade. Nevertheless dangers remain intwo interrelated areas: macroeconomic stability, and public finance management. Without improvements inthese areas, the country's high growth rate will be put at risk. xxxii. Macroeconomic stability. Today it i s common cause that macroeconomic stability and predictability i s the crucial foundation for growth and poverty reduction. This encompasses bothprice stability andfiscal control. The immediate losers from an inflation spike are the poor, as it acts as a regressive tax. Inthe medium term the poor lose also, because investors are scared away and growth and employment opportunities decline. xxxiii. Despite the country's growth record and progress on the macroeconomic front, there i s still a danger o f price instability. Reorganizationo f the banking system inthe mid-l990s, together with tight money, resulted in single-digit inflation up to 1999. But there have been two phases o f inflation brought on by lax monetary policy, one starting in 2000 and resultingininflation o f 17percent in2002, and another in200314 resulting in inflation o f 17 percent inearly 2004, subsequently falling to 8 percent by January 2005. Inthese instances, monetarypolicy hadto betightenedby increasingreserve requirements, raising the bank rate andor accelerating the sale o f foreign exchange reserves to the public. xvii COUNTRY ECONOMIC MEMORANDUM xxxiv. By less developed country standards, Mozambique has maintained a good record o f fiscal discipline inthat it has on only a few occasions resortedto borrowing from the banking system inorder to make up for a fiscal deficit.' The public internal debt stock i s around 4 percent o f GDP. xxxv. The present problem on the fiscal side i s that revenueperformance slackened during2004, andit needs to improve because the Government's spendingrequirements in education, health and infrastructurewill likely continue to rise, at leastpari passu with population. Inaddition, aid as a percentage o f GDP i s likely to fall from its very high level o f 15 percent currently. The plan o f the PARPA was to raise revenue to 15.4 percent o f GDP by 2005; by 2004 this had reached only 12 percent. Hence it i s imperative that Mozambique raise more domestic revenue to support its poverty reduction programs and ultimately reduce aid dependency. xxxvi. The public external debt was considerablyreduced through two HIPC operations. Combined with the country's excellent prospects for exports, the net present value o f foreign debt service divided by exports i s projectedto fall from its 2004 level o f 84 percent to 50 percent by 2020. Values o f 125-150 percent are considered inthe HIPC context to indicate a debt overhang problem. Hence by this standard, Mozambique is well clear o f a growth-reducing debt overhang, at least on the external side. xxxvii. Mozambique's exchange system i s a managed float. The real effective exchange rate has slightly depreciated since the 1990s. Even though aid payments have amounted to 12-19 percent o f GDP, this has not caused a significant appreciation o f the currency. A general equilibrium model built to test this indicates that the increase in productivity generated by the aid moneys (e.g. roads, education) has compensated for the aid-driven upwardpressure on prices o f nontradables, thereby sparing the country from the export-depressing effect o f "dutch disease". The key point i s that the exchange rate should continue to be responsive to market trends. xxxviii. Publicpnancemanagement. Mozambique's fiduciary accountability can at best be described as weak, although the Government i s taking measures to improve it. The weakest areas are accounting, auditing and procurement. This is a matter o f concern inits own right. Itwill also be criticalfor growthinthe next decade, because: P improved spending efficiency is neededto achieve better service delivery, which inturnis neededto spur growth, particularly ininfrastructure, education and health. It has been found that public works are substantially (20-30 percent) more expensive in Mozambique than inthe region; P total expenditure can no longer expand at the rate it usedto. Total expenditures are likely to remain at 24-25 percent o f GDP. Hence it will be essential to secure efficiency improvements inthe future; P without steady improvement inpublic finance management some o fthe donor fundingmay be put at risk. This is important because Mozambiqueis receiving more aid thanmany other countries. The justification for such generosity i s that the aid i s being well used - as evidenced in highgrowth rates andbig 'A notable blemish on this record occurred in2001-2 when two banks were bailed out. The country is still bearing the cost o f the "inefficient forbearance" that led to the banking crisis, because the bonds have not yet been repurchased and the issuance o f new paper i s a contributory factor to highinterest rates. xviii EXECUTIVE SUMMARY improvements insocial indicators -but if any o f these i s let slip, Mozambique's specialjustification will fall away and its rather weak fiduciary record may well result inits aid levels falling. xxxix. The Government has embarked on a sequenced program o f reform, starting with a new financial management law in2001 and accompanyingregulations in2002, which set the basis both for modem accountingprocedures and for procurement reform. The electronic financial management system SISTAFEi s beingpilotedintwo ministries in 2005. Next will be the rollout to other sectoral ministries during2006. Then should follow the improvement o f the classifiers (program, functional, regional) so as to provide detail sufficient to permit analysis o fpoverty impacts. Inparallel, the regulations and institutions goveming procurement are beingrevisedso as to promote competition and transparency. xl. A further important task i s to coordinate the SISTAFEimplementation with the reform o f public sector, and inparticular to enable coincidence between the staff establishment heldby the Ministerio da Administrag50 Estatal and the pay list o f the Ministry o f Finance. Inso doing it will be necessary to eliminate supemumerary staff: the Government's public expenditure review of education in2003 concluded that as many as 20 percent o f the staff inprimary educationwere not properly accounted for.8 Challengesto growth and poverty reduction:privatesector development xli. Here we refer to the long pendingagenda o f reforms aimed at promoting private sector initiative inbusiness and removing barriers to investment by domestic and foreign firms. There i s little inthis section that i s distinctively new. The contribution o f the Country Economic Memorandum inthe area o f private sector development has been to come up with a feasible prioritization of tasks - legislative, regulatory, administrative - to facilitate businessinvestment. This prioritization has beenconcisely summarized in Table 28 (p. 68). Inthe following paragraphs, a "short list" o f interventions are discussed: > facilitating trade > energy reform andinvestment P financial access xlii. Facilitating trade. Rapid expansiono f exports has been an important source o f growth during the past 10 years inMozambique. Since the early 199Os, exports have expanded at arate (10 percent) substantially higher than world exports (6 percent). Mozambique i s one o f the few countries inAfrica that managedto increase its share in world exports. This rapid export expansion was ledby private initiative, both national and foreign. xliii. As inother countries, for Mozambique, aprivate sector-led, export-based growth strategy hast the greatest long-term potential to sustain a high growth rate inlight of the existing conditions: the domestic market i s limited, and has relatively low purchasingpower. Additionally, giventhat the national saving rate i s not highenough to generate adequate growth to make a marked impact on poverty, a particular emphasis needs to be placed on attracting foreign direct investment. * See the discussionon educationsector employment inWorld Bank (2003b). xix COUNTRYECONOMICMEMORANDUM xliv. Among the reasons for Mozambique's export success i s persistence inreform. Mozambique has liberalized its imports significantly, as was noted above (paragraph xii, page xii). Mozambique i s committed to further liberalization through SADC Trade Protocol, and will negotiate further duty phase down inthe context o f the Economic Partnership Agreement (EPA) with the EU. xlv. M a n y challenges remain. It was noted above (paragraph xii, page xii) that most export growth was due to mega-projects, and that "traditional'' exports grew slowly. While the production o f aluminum, gas, titanium and electricity will continue to be the main sources o f export growth, the key challenge inbroadening the base for value added growth to achieve poverty alleviating objectives will be the removal o f barriers so as to allow small and medium enterprises to contribute more. xlvi. M u c h needs to be done to improve the support services for exports to eliminate the anti-export bias inherent in the import duty system. The priorities are: extension o f export incentives to small companies, acceleration o f value added tax (VAT) reimbursement, improving the access o f firms to the financial system, and improving standards and the capacity to meet the sanitary and photo-sanitary requirements, particularly for the EUmarket. Given that the macro-economic and fiscal conditions are improving and the exchange rate regime i s broadly appropriate, the key constraints on promoting export-led growth are "behind-the-border issues": > the weak infrastructure system (including electricity and roads), which still requires enormous investment as well as policy improvement, and > the cumbersome investment climate (e.g. slow import clearance, complex taxation procedures, slow business startup, a non-functioning judicial system, high transportationcosts, and highlending risk). xlvii. Energy reform and investment. An area calling for strong policy orientation, as well as substantial investment, i s energy. Some 64 percent of manufacturing firms rate electricity as their most serious infrastructure problem. They report some 17.5 power outages per month. Electricity access increased from 4 percent in 1996/7 to 7 percentgin 2002/3 (240,000 households), but the bulk o f the new connections were inthe best-off quintile o f households, whose access rose from 11percent to 18 percent. Inany case, the pace o f new connections - at the historical record o f 8,000 per year" - i s only slightly higher than that of population, so that the l o w electrification rate is set to stagnate. A s things stand, the poorest 80 percent o fhouseholds will not obtain electricity for decades. This envisaged future stagnation inelectricity access contrasts unfavorably with the improvements that are foreseen intelecoms, water, roads and other aspects o f infrastructure. xlviii. An integratedand coordinated approach is called for, combining policy, restructuring and increased investment. Inthe short term, > The financial and operational performance o f Electricidade de Moqambique (EdM) shouldbe upgraded, and energy losses reducedfrom 20 to 8-10percent. See Fox et al. (2005), Table 5. lo deBarros(2005),paragraph52. Between 1996and2000(4years), therewere 31,000new See connections. xx EXECUTIVE SUMMARY k Consumer services shouldbeimproveddrastically, for example, bygreatly expanding pre-paid services, thereby improvingrevenue collection. k Investment shouldbeincreasedto intensify connections where thebackbone supplylines are already inplace. The network shouldbe further rehabilitated and reinforced. Inthe mediumterm, k Institutionalandlegalreforms shouldaimat establishing anenablingenvironment conducive to the sector's development, with adequate regulatory capacity. xlix. Improving access to and the cost offlnance. The vast majority o f firms in Mozambique state that the lack o f access to finance, and its highcost, are severe problems for their growth. Interest rates are higher inMozambique than incountries at a similar level o f development. This i s partly because interest rate spreads are high(19 percentage points), inturnbecause o f highloan losses and highoverhead costs. Only 29 percent o f firms have bank loans; several countries at similar income levels do much better inthis regard, e.g. Eritrea at 45 percent o f firms. Small firms face even greater barriers interms o f access. 1. Several reforms are therefore indicated. Banking supervision needs to be strengthened further, through bringing loan classification into line with international practise andbuilding the core knowledge o f supervision staff. A commercial court could be established, dealing with the largest cases. Simultaneously, morejudges shouldbe trained incommercial dispute resolution, andjudicial procedure should be simplified to reduce procedural complexity. The scope and reliability o f the credit registry administeredby the Bank o f Mozambique needs to be increased. To develop the capital market, which i s very small, the first objective should be to develop further the market for public securities, thereby providing a market-driven benchmark for the issuance o f private securities. li. Finally, improving access to finance for micro enterprises would also contribute to lessening the constraint imposed by the highcost o f capital inMozambique. Although the microfinance industryhas grown, its outreach i s small, with a high concentration inMaputo. Important bottlenecksneed to be eliminated, including restrictions on deposit taking, The current legislation under preparation should allow micro finance institutions to take deposits. It would be well served ifit benefited from expert advice in finalizing the micro finance law. Challengesto growth and povertyreduction:naturalresources lii. An important part o f increasing growth and reducingpoverty i s improving the regulatoryframework for natural resources, specifically inwater, mining, rural land, fisheries and forestry. These areas receive this focus, and are gathered together inthis section, because (a) collectively they are vital for economic growth, (b) they are key for poverty reduction (especially water) and for subsistence (forestry, fisheries, rural land), (c) their misuse would result inenvironmental unsustainability and would threaten the growth process as a whole, and (d) certain o f them serve as excellent case studies for efficiency, governance and investment perspectives. liii. InMozambique, certain characteristics are commonto these five natural resource areas: x x i COUNTRYECONOMIC MEMORANDUM 9 anoverall regulatoryenvironmentthat seeksto protectthepoor through open access andor zero rent extraction by the state; 9 areluctance to usetransparentprocedures for allocationofresourceuse rights; 9 under-collectionofrentsthat would rightlyaccrueto the state, and from this follows an inefficient use o f resources, less economic growth, lower government revenues, and consequently less poverty reduction. Hence, as will be explained inthe ensuing paragraphs, improvements in natural resource management will be crucial for growth andpoverty reduction. liv. Water. Of the five natural resources, the one needing the most attention by policy-makers i s water: small-scale irrigation, flood and drought mitigation, and water supply all have roles to play inpoverty reduction. lv. Waterfor irrigation: About one-tenth o f the country's 36 million hectares o f cultivable land i s suitable for irrigation. Land under cultivation i s 4.3 million hectares, o f which only 0.9 percent i s actually irrigated- the least inthe region. lvi. Ina context of limitedwater storage capacity, and a lack o f financial resources, the best use o f public resources would be to concentrate on improving conditions for smallholder farming. For cost effectiveness, the focus would best be on small-scale, low- cost schemes rather than large projects. Encouraging smallholder irrigation would also lead to increased high-value crops and increased incomes for some o f the nation's poorest households. While there i s substantial potential for the expansion o f commercial irrigation, this should be driven by the private sector. The role o f the Government i s to put inplace the legal, institutional andregulatory framework, as well as to support the development o f water storage infrastructure providedthis i s linked to multi-purpose projects (energy i water / transport i drought management). lvii. Water resources infrastructure. The useful capacity o f Mozambique's dams represents only 5 percent o f the mean annual runoff o f the country's rivers, excluding the Zambezi. Thus the country needs to develop its infrastructure to store a larger share o f the runoff. lviii. The Government could make significant progress inreducing vulnerability to droughts by dealing with issues related to rural water supply, irrigation water for smallholder farming, and the operation andmaintenance o f existing systems. This could be complemented by negotiation with upstreamriparians (to ensure that the operation o f upstream reservoirs does not aggravate floods inMozambican territory), and waming systems that would serve both for the management o f droughts as well as floods. Small dams and reservoirs should be built inthe most drought-prone areas. Two major dams are urgently needed for the water supply o f Maputo and o f Beira and for irrigation inthe Lower Pungoe. Medium sized dams are also required to ensure the water supplyto several lesser cities. For flood attenuation, storage reservoirs are called for, provided the operating rules for all new and existing dams are reviewed and adjusted to balance needs interms offlood attenuationandother uses. Major beneficiaries o fthis infrastructure would be the poor, who tend to be more vulnerable to meteorological risk. lix. Water supply. The lack o f clean water i s a factor inpoor health and lowered work productivity. Here we build on work done by the Public Expenditure Review o f 2003 -informed by the dramatic finding, which was not available for the latter study, that rural water access i s lower than had been thought (namely 27 percent as opposed to xxii EXECUTNE SUMMARY 40) and urban water access is higher than hadbeen thought (64 percent as opposed to 38)." M u c h improvement i s called for, particularly inrural areas. lx. Recent new directions inurbanwater supplyhave turned out to be a success story. As o f 2002, household connection coverage inMaputo amounted to 70,000, and water ran only 9 hours per day. The government opted for a "delegated management" model, introduced by competitive biddingfor five major cities. A private sector contractor commenced operations in2002, Within two years it was evident that the right decision had been made: coverage inMaputo increasedto 90,000 connections, and supply improvedto 13 hours, and the urbanwater sector has attracted further investment. lxi. Inparallel, the government has pursuedthe "demand driven approach" inrural areas, which requires communities to make a specific request for the installation o f a water source, pay an upfront fee, and take responsibility for its maintenance. Even so, at any one time about 30 percent o f wells are non-fimctional for lack o f spare parts. lxii. Despite the evident progress made, persisting with the present arrangement i s inadequate. Whereas in access to basic health and education Mozambique frequently equals or exceeds the Africa average, inrural clean water access it i s well behind, at 27 percent, compared with Africa's 46.5 percent. Gross primary school enrolment i s now over 100percent, and vaccination coverage 82 percent, so the "missing link" inrural livelihoods i s now the absence o f clean water. lxiii. An integratedstrategy o fswift expansion of access to clean water should maintain two principles: cost-effectiveness, and the "demand-led approach" - for absent the latter, the proportion o f non-functional wells will quickly soar beyond its present 35 percent level. Inaddition, mechanisms should be designed to ensure that spare parts are available as required and that community capacity i s built up. This mammoth expansion can be attained by Increasing resources and personnel inrural water access. This may entail reallocation from sectors o f lesser priority. Although the approach will continue to be demand-driven, the bindingconstraint presently i s not communities that satisfy the criteria but staff, capacity-building and fundingto permit the roll-out. Investigate thefactors underlying high drilling costs. Inparticular, consider whether subjecting more work to tender and having less done by parastatals would reduce costs. Expand standpipe access inurban areas. This can be done at lower cost than by expanding household tap connections. One possible management model i s subcontracting to individuals (as i s beingdone in some Mozambican towns such as Angoche). Licensingprivately operated water kiosks as i s done inother African cities. Increase co-payment in urban areas. Inall cities, and particularly the smaller ones, the tariffs need to be raised so as to cover operation and maintenance and make a contribution to debt service. "Figuresfromthehouseholdsurveys. Accesstowaterfromimprovedsourcesinruralareaswasraised from 12 percent in 199617 to 27 percent in 2002!3, and in urban areas from 56 percent in 199617 to 64 percent in200213. See the discussion inparagraph 270 (p. 80) and footnote 128 (p. 80) about measurement o f access to clean water. xxiii COUNTRY ECONOMIC MEMORANDUM Zxiv. Mining is becoming an increasingly valuable naturalresource. The recent purchase by a Brazilian mining firm o f exploration rights inthe Moatize coalfield has underscored the excellent prospects for this sector. Presently the sector accounts for just under half a percent o f GDP, and contributes fiscal revenues o f $3-5 million (apart from the payments by the latter coal prospecting company). The number o f firms has doubled since 2001. This i s partly due to progressive legislative changes and the introduction o f an innovative and transparent MiningCadastre, which incidentally serves as an example for other needed institutional reforms (e.g. land, fisheries, forestry). Artisanal and small- scale miningprovides an occupation, full-time or part-time, for 100,000 people scattered throughout the country. lxv. However, Mozambique's miningoutput i s still well below that o f countries with similargeological potential. Annual exploration expenditures could easily rise from their present level o f US$15-25 to US$ 50-100 millionper year. lxvi. To achieve this, the system o fminingtaxation needs to be brought inline with the tax arrangements for the remainder o fthe economy. The current system is globally uncompetitive. It consists o f the general income tax code (Impost0 sobre o vendimento daspessoas colectivas -IRPC), plus the royalty and the surface tax for the holding o f permits, plus the Cddigo dos Beneficios Fiscais (2002). The system i s unattractive to private investors because the royalties can place a large burden o n companies during mineral market downturns, and because the IRPC does not permit accumulation and carry forward o f expenditures during the exploration and development phases. Hence the arrangements for miningfirms need to be revised, to make them globally competitive, and to secure a fair share for the government. lxvii. Artisanal miners, who are generally poor, should continue to enjoy free access to mineral resources. lxviii. Rural land administration has been reformed by a new land law (1997) and regulations (1999): customary tenure arrangements gained official sanction, and consultation was mandated prior to issuance o fnew concessions. The law succeeded in its main aim o f protecting the traditional land use rights o f smallholders. lxix. These reforms are still inadequate. Since land cannot be owned, the only barrier to acquiring rights to its use i s bureaucratic; not surprisingly, concessions to large tracts o f land are still being awarded virtually gratis. The lack o f an official landmarket creates opportunities for rent seeking and is inefficient because land rights will not naturally gravitate to their most productive uses. When improvements inrural areas are sold, the land concession on which they stand does not automatically accompany the sale. Further permissions have to be secured before the concession may be transferred, creating further opportunity for rent-seekingbehavior. A further difficulty is that concessions are subject to bureaucratic interference, due to the requirement for a landuse plan. Such interference creates ajustified perception on the part o f potential large-scale farmers that tenure i s too insecure to merit large investments. lxx. Substantial efficiency improvements could be obtained by P changingthe regulations to allow landuse rightsto automatically accompany sales o fbuildings and improvements, without bureaucratic interference; and P substantially increasingthe landtax for medium- and large-scale concessionaires, thereby obliging landholders who are not making productive use o f their land to xxiv EXECUTIVESUMMARY transfer it to others who would. The concessionaires couldbe compensated by being relieved o f the requirement for use plans altogether. This would increase the perception o f land tenure security and encourage further investment. lxxi. Fisheries account for 1.5 percent o f GDP and 10-15 percent o f exports; most o f this arises from industrial and semi-industrial shrimpfishing. Artisanal and subsistence fishingare also important inthat they provide employment for an estimated 90,000 people. There are three large industrial and some 200 semi-industrial companies involved in shrimp fishing. License fees amount to US$5 million annually, although this falls well short o f the full resource rent for reasons which are elucidated inthe main text (see 304, p. 88). The Government has been pursuing a strategy o f promoting local (viz. Mozambican national) participation inshrimp fishing, usinga Government-controlled firm Emopescato formjoint ventures. This was achieved by issuing substantial numbers o f new quotas to new entrants, and this without reducing the industrial fleet participation. Not surprisingly, the total fishing effort has increased by a factor o f three since the late 1970s, although the total catch has remained nearly constant.'2 lxxii. However, the allocation procedure o f the new licenses was not transparent, with substantial values being transferred without clear regulation or distribution criteria. In any case the arrangement has not been as successful as had been planned at the start, in that the Mozambican operators have not been able to develop their share and national operators remain scarce. Ixxiii. The leakage o f the resource rent needs to be addressed. This could be achieved by a "two-track" arrangement. International accesx could be arranged best by an auction system, This would considerably increase the total rent receivedby the Mozambican state.13 The reason for not advocating the "first-best" efficient method o f an auction system throughout i s that this would likely exclude national operators almost altogether. Hence for national access, a clear set o f publicly available criteria for quota distribution i s required, to ensure predictable and fair systems for users, and also to secure social benefits from natural resource exploitation. Such "grandfathering" should be phased out within a predefinedperiod of, say, five years. lxxiv. Artisanal and subsistence fisheries are an important poverty buffer. General community development has been initiated as part o f fisheries development, with assistance from the Instituto de Desenvolvimento Pesqueiro de Pequena Escala. Artisanal fisheries are subject to regulation, and it i s not proposed to change this. Subsistence fisheries enjoy open access, which is the best approachwith this fishery; it would be important not to undermine the poverty buffer through improper management or other governmental interventions. There i s no indication that open access inthis case has been leading to depletion o f the resource. lxxv. Forestry accounts for 2.5 percent o f GDP presently. The output o f the sub- sector i s about 120,000 cubic meters per year. Experts put the potential at some 500,000 cubic meters. Major policy changes have recently been instituted. The logging fee was sharply raised, to the point that total fees rose from US$0.5 million in2002 to US$2.3 million in 2003. A further doubling o f fee levels is planned for 2005. Plausible l2Sincethereisclosedseasonregulationinplace, thereis stillnodanger ofresource depletionatcurrent levels of effort. l3Seethe maintext for the data. xxv COUNTRYECONOMIC MEMORANDUM computations indicate that these resultant fee levels will approximate the scarcity rent. Another major reform is associated with the authorities' efforts to create greater value added locally. With the aim o f encouraging local wood processing - as opposed to the export o f uncut logs - the authorities imposed an export ban on high-quality uncut wood in2003. lxxvi. A disadvantage withthe current system o f fees ("volume license fee") is that it differentiates by wood type, which i s appropriate, but not by locality. This gives undue incentives to harvest wood near populous areas, and discourages harvesting indistant areas. This i s an inefficient arrangement, causing under-use inremote areas where there are no ecological problems, and threatening deforestation inareas o f easy access. lxxvii. Inthe mediumterm, a more effective way o fbringingabout increased competition would be a bidding license regime. This way, instead o f the government's setting the fees per log, the loggers would themselves (in effect) propose fee levels, and because they will be doing this competitively, there would be a substantial increase o f revenues - possibly as much as $1.5 m. lxxviii. The export ban could also be reexamined. It i s not clear that this i s the best way to accomplish the objective o f enhancing local value added. For one thing, the existing plants are as yet unable to process all the wood available, andnot surprisinglythe total amount o f wood produced has fallen. Inaddition, there i s a loss o f potential revenue, if only because the ban i s not effectively policed. Most importantly, the link between the export ban and the potential investors i s weak and so this measure i s unlikely to have the desired effect. More effective, and fiscally advantageous, would be a carefully graduated combination o f time-bound tax and licensing advantages, a time-bound and light export tax, and infrastructure incentives. lxxix. The forestry sector i s also important as a contributor to the livelihood of tens o f thousands o f rural dwellers. For this reason, the present arrangement o f issuingsimple licenses over the counter for 500 cubic meters o f cutting annually should be maintained. Growth and povertyoutlook lxxx. What o f Mozambique's growth performance inthe fbture? A general equilibrium model suggests an average long-term growth rate o f 7 percent, from which should be subtracted 1percent per year owing to HIV-AIDS (see paragraphxxvii, page xvi) and another 1percent per year to take account o f meteorological risk, arriving at a long-termgrowth rate of 5 percent. This is better than the region's 3.4 percent- appropriately due to Mozambique's superior natural resources. It i s poorer than Mozambique's recent past growth o f 8 percent. It would be hazardous, however, to assume a "business as usual" rate o f 8 percent, because 9 such "growth spurts" rarely extendmorethan a few years, 9 Mozambiquefaces meteorologicalandother risk, 9 the agricultural catch-up isnow exhausted andthe sector is growing at a lesser rate than it did before, 9 most ofthe "easy" growth-enhancing reformshavebeendone, e.g. telecoms liberalization. This leaves the more challenging institutional reforms such as judicial reform, and xxvi EXECUTIVE SUMMARY P aidwill fall as a share o f GDP. lxxxi. With a vigorous reform agenda, onthe other hand, it may bepossible to attain growth rates o f 7-8 percent inthe medium to long term. A "virtuous cycle" could be set inmotion, as the authorities pursue institutional upgrading with what mightbe termed "second-generation reforms". So as to convince the skeptic that this i s possible, four examples o f far-reaching - and sometimes painful - institutional change are cited: P Reforms inair transport in2001 ledto competitive new entry of a domestic operator, and to increased tourism using charters. P The liberalization o ftelecommunications in2001 permitted avast increase: from 85,000 land lines and 51,000 cellphones, to 700,000 cellphones by 2004, with better service and the lowest prices inthe region. 9 Adoption ofatransparent andnon-discretionaryMining Cadastre,together with a new pro-investment mininglaw, resultedina sharp increase inapplications for miningtitles, from 110in2001 to 180in2004, and the time taken to obtain a license fell from 120days in 1999 to 50 days in2003. The groundwork has been laid for long-term growth inthe sector. P As noted above (paragraph Zx, page xxiii), the "delegated management" ofurban water led to important improvements in coverage and supply times. These experiences could be repeated inother sectors, along the lines suggested inthe "Challenges" above. Zxxxii. Simulations using the "growth regression" literature suggest that Mozambique could enhance its growth performance by improving its business climate to East Asian levels, Gains o f up to 1percentage point per annum could be made by avoiding banking crises, or by improving governance, or by reducing government nonproductive spending. Bigger gains o f up to 3 percentage points per annum could be secured by - intime - improving infrastructure, or by raising access to secondary education. lxxxiii. Provided the country's growth momentum i s retained, and provided inequality does not increase much, poverty rates will continue to fall. W e assume GDP growth o f 5 percent a n n ~ a l l y . 'Then from 54 percent in 200213, the povertyheadcountmay fall as ~ far as 40 percent by 2007. Andthere is a possibility o f accomplishing the Millennium Development Goal o f halvingthe 1990 poverty rate by 2015. (If,on the other hand, inequality increases sharply as it has inmost fast-growing countries, then poverty will fall less slowly.) Summary of recommendations lxxxiv. The full set o frecommendations are given inthe main text. Here, for the sake o f clarity, the key recommendations are highlighted. W The first and most important is macroeconomic management: P Macroeconomicstability. Without low inflationand fiscal control, none o fthe structural reforms will succeed inmaintaining good growth rates and reducing poverty. Given the Government's plans to expand poverty-related spending, the l4 Growth rates are higher for 2003-2004, and also for the short-term projectionperiod 2005-2006. Thereafter the long-term expected growth rate of 5 percent kicks in. xxvii COUNTRYECONOMICMEMORANDUM only way that crippling deficits canbe avoided is through raising revenues by broadening the tax base and creating an independent tax agency. P Publicfinance management needs dramatic upgrading because Mozambique is among the poorer performers in Africa and without substantial improvement, donor largesse which has supported the country's growth will bejeopardized. The SISTAFE should be implementedthoroughly, "off-budgets" brought on budget, and budget execution figures reported meaningfully. W Inthe area o fprivate sector development: electricity reliability shouldbe improved, and access to the network increased at a faster rate, through restructuring of EdM. W Inthe area o fnaturalresources: water access for rural people should be greatly expanded by reallocating funding from sectors o f lesser priority. Zxxlcv. T w o further recommendations emerged from the analysis. W It is essential that the Government persist with its strategy for combating HIV/AIDS. Mozambique is one o fAfrica's hardest hit countries, with a prevalencerate of 16.2percent. The cost to the economy i s o f the order of 1 percentage point o f GDP per capita each year. W Growth inagriculture has started to plateau out after the post-conflict "bounce- back". There will be a needto move to productivity enhancingmethods. These will likely include improved farmer-relevant research, broader dissemination o f drought-resistant varieties, the encouragement o f outgrower arrangements, the exploration o f new models for access to financial markets by smallholders, facilitating investment by medium- and large-scale farming, and upgrading rural infrastructure. xxviii SLWARIO EXECUTIVO SUMARIOEXECUTIVO IntroduqBo Ixxxvi. Moqambique conheceu umperiod0 de recuperaqgo dramaticada destruiggo provocada pela guerra civil. Desde 1992, a infra-estrutura foi melhorada e esta agora a aproximar-se dos niveis antes da guerra e os rendimentostem estado a registar um aumento consideravel. 0 indice de pobrezareduziu de 69 porcento, em 1996/7, para 54 porcento, em 2002/3. Entre 1996 e 2003, a economiaregistou um crescimento medio anual de 8 porcento. Este sucesso pode ser atribuido a abordagem faseada mas determinada do Govern0 de estabilizaqgo e reformas estruturais, bemcomo iassistencia concessional (metade das despesas do Govemo), uma extraordinariarecuperaqgo agricola, expansgo das exportaqaes agricolas e uma rapida expansgo do turismo, da construqgo e de certos subsectores de manufactura. Umoutro factor tambCm significativo foi o Cxito das autoridades na atracqgo de "megaprojectos" nas areas de fundig80 de aluminio, gas natural, e mineraqgo de titinio, o que levou a que as exportagdes fossem triplicadas num espaqo de trks anos. lxxxvii. Outro factor que contribuiu para este sucesso foi o facto do pais ter conseguido alcangar a reconciliaqzo, pondo fim iguerra civil e gerindo os conflitos potenciais desde entgo. Moqambique acaba de realizar as suas terceiras eleiqaes gerais e presidenciais. lxxxviii. N o entanto, o pais continua pobre (com rendimento per capita de 240 dolares); a infra-estrutura e inadequada, as necessidadesem termos de educaq8o e saude ainda n8o foram satisfeitas e os indices de pobreza continuam elevados. Muitas das reformas de "primeira geragso" associadas com a liberalizaqgo de mercadoj a foram implementadas. 0 pais enfrenta agora a possibilidade de maiores constrangimentos macroeconomicos, a necessidade de melhorias institucionais substanciais de modo a tomar o crescimento sustentavel, uma necessidade crescente de uma melhor priorizaggo e gestiio das despesas publicas, com vista a erradicar a pobreza absoluta, e grandes investimentos em infra-estrutura para promover o crescimento e ter ummaior impact0 na reduq8o da pobreza. lxxxix. Objectivo e estrutura do Memorando Econo'micodo Pais. 0 presente Memorando analisou o binomio crescimento -pobreza, usando uma grande variedade de fontes de dados, incluindo o recentemente concluido levantamento familiar nacional (2002/3). Procurou compreender as fontes de crescimento no passado recente, avaliar as perspectivas de crescimento na proxima decada, examinar as provaveis implicaqdes na pobreza e delinear as politicas que sera necessaria introduzir para promover um maior crescimento e lograr a reduqgo da pobreza. A razHo deste foco neste momento prende-se com o facto do Govemo estar a planificar o desenvolvimento do seu segundo Plano de AcgSio para a Reduqgo da Pobreza Absoluta (PARPA) e sentiu-se que umenfoque maior no crescimento e pobreza contribuiria de forma substancial a este processo. xc. Alemdisso -emconsequhciada coberturasuperficial de que o assunto foi object0 em outros trabalhos do sector, produzidos ou n8o pel0 Banco - o Memorando Economico do Pais tambkm analisa a relevincia da gestiio dos recursos naturais aos objectivos relacionados com crescimento e pobreza. Foi constatado que os recursos naturais assumiam uma grande importincia tanto no crescimento como na reduq8o da pobreza. Assim, espera-se que futuramente as questdes relativas aos recursosnaturais xxix MEMORANDO ECONOMICOPAk DO sejam totalmente integradas no debate das politicas nacionais. xci. Depois do inicio da preparaqiio do Memorando do Pais, o Govern0 fez umpedido adicional, que este incluisse umcapitulo sobre o desenvolvimento do sector privado. Este aspect0 niio constituia partedo plano inicial, pois, muitapesquisa valiosaja tinha sido realizada sobre esta questiio importante (conforme e documentado na pagina 47). Assim ficou acordado que a contribuiqtio do Memorando niio envolveria a introduqtio de novos desenvolvimentosna area do sector privado, mas sim produziria uma sintese e priorizaqgo exequiveis das diferentes recomendaqaesque resultam destes esforqos. xcii. 0presente Sumario Executivo agregaos argumentos nas seguintes secqdes: 3 Trajectoria do crescimento e pobreza 3 Desafios ao crescimento e areduqiio dapobreza: desenvolvimento do sector privado P Desafios ao crescimento e areduqiio depobreza: recursos naturais P Previsiio do crescimento e situaqiio depobreza Trajectoria do crescimento e pobreza xciii. Moqambique tem presentemente uma populaqzo estimada de 19 milhdes de habitantes e as previsdes apontam paraum crescimento anual de 2.4 porcento durante o period0 2005-2010. Dado que as taxas de dependencia estZo a baixar, a dinamica demografica niio e um factor preponderante nas tendencias da pobreza; com efeito, a dinimica demografica tem contribuido para o aumento dos rendimentos per capita e para o declinio nos indices de pobreza. Espera-se que, ate 2010, a populaqiio urbana activa registe um crescimento de 4 porcento, sublinhando a necessidade de uma via de crescimento que privilegiea criaqHo de oportunidades de geraqgo de emprego. xciv. Comofoi conseguido o crescimento de Sporcento? As condiqaes para um crescimento sadio foram conseguidas graqas a um ambiente macro estavel, avanqos na liberalizaqiio dos sectores-chave @or exemplo, telecomunicaqdes, transporte aereo) e progressos na liberalizaqiio do comircio intemo e externo. Alem disso, ocorreram varios factores extraordinarios, incluindo a ajuda concessional substancial. Estes factores siio discutidos nos paragrafos que se seguem. xcv. Relativamenteao ambiente macro, a inflaqiio baixou de mais de 60 porcento para niveis de um digito nos finais da decada de noventa e, nZo obstante o resvalamento que se seguiu a crise da banca em 2000-2002, espera-se que no futuro permaneqa nos niveis de umdigito. Os gastos do govern0 aumentaram de forma significativa (24 porcento do PIB, em 1997, para 30 porcento, em 2002, baixando posteriormente para 24 porcento em 2004), mas contraiu poucos emprestimos do sistema bancario, pois, os gastos foram financiados atraves do aumento da ajuda concessional (e de um aumento moderado das receitas). Desde 2001, o Govemo prossegue uma orientaqgo fiscal e monetaria guiada pel0Plano de Acqzo de Reduqzo da Pobreza Absoluta (PARPA), que releva o papel- chave da estabilidade macro, prevk um aumento progressivo das receitas e restringe a tomada de emprestimos do sistema bancario. xcvi. Relativamentea liberalizaqzo, a intervenqiio do Estado na economia agraria foi quase suprimida em meados da dkcada de noventa. As telecomunicaqdes t&msido liberalizadas de forma significativa a partir de 2001 e os aparelhos de telefonia move1 aumentaram de 51.000, em 2001, para 700.000 em 2004. 0 trafego aereo foi liberalizado xxx SUMARIO EXECUTIVO em 2002 e 2004 viu o primeiro novo concorrente aoperar nas rotas domCsticasbem como os primeiros v6os turisticos. xcvii. Quanto ao comercio, durante a dCcada de noventa, as tarifas de importaqgo foram gradualmente reduzidas at6 atingir umatarifa media de 9 porcento, uma das mais baixas emAfrica. Preve-seque, ate 2006, ataxa tarifaria mais alta se situe naordem dos 20 porcento. A gestgo da alfindega foi subcontratada durante a dCcada de noventa e a eficiencia da colecta tributaria melhorada. As exportaqbes tbm registado um forte desempenho, crescendo a uma taxa anual de 22 porcento (emtermos de USD),bem acima da media das exportaqbes mundiais (6 porcento), tomando Moqambique num dos poucos paises africanos cuja fatia mundial em termos de exportaqbes aumentou. Contudo, a maior parte da expansgo das exportaqbes foi graqas aos megaprojectos; tendo as exportaqbes tradicionais registado umcrescimento de apenas 2,3 porcento. xcviii. Alem disso, Moqambique beneficiou de tr&s factores excepcionais: P A agricultura, que 6 abasede subsistencia de 80 porcento dapopulaqgo, registou um crescimento de 6,8 porcento por an0 entre 1992 e 1997, que abrandou para 4,6 porcento anuais entre 1997 e 2003 (incluindo o impacto das inundaqaes de 2000). Os principais contribuintes sgo os cereais (milho), aqucar e tabaco. Umaparte do crescimento registado entre 1992 e 2003 foi devido h expansgo de area (2,4 porcento) e ao aumento da forqa de trabalho, e o resto foi devida a melhorias dos rendimentos. 0 crescimento em certas regibes como Tete foi mais forte devido ao comhcio, por exemplo, o comCrcio transfronteiriqo de milho e de outros produtos com o Malaui. Este factor e classificado como "excepcional" porque uma parte substancial do crescimento registado foi devida a "recuperaqgo" depois da guerra. E uma recuperaqgo em dois sentidos -parte do crescimento agricola verificou-se em forma de rendimento naproduqgo de culturas, que tinha declinado acentuadamente durante o periodo socialista e de guerra, recuperando rapidamente aos seus niveis iniciais durante a d6cada de noventa. Num outro sentido, durante a dCcada de noventa, a area cultivada aumentou aos niveis conseguidos antes da guerra, com o retomo dos refugiados de guerra aos seus campos agricolas. AlCm disso, os rendimentos na"o agricolas (microempresas, salarios e remessasde valores) foram notaveis nas tres provincias mais meridionais que dependem das ligaqbes com a cidade de Maputo. > Moqambique e um dos maiores recipientes da ajuda concessional em Africa (cerca de 12-15 porcento do PIB, representando metade dos gastos do. Govemo). A ajuda contribuiu significativamente aos esforqos do Govemo na consecuqgo dos seus objectivos de melhorar o acesso e a qualidade dos serviqos - ajulgar pelas melhoras na educaqgo, saude e vias de acesso -e na relativa melhoria do nivel de responsabilidade fiduciaria. (A excepqgo 6 o HIV/SIDAque continua propagar-se). As abordagens de modelaqgo demonstram que a ajuda respondeu por uma parte substancial do extraordinario crescimento registado em Moqambique. P Umoutro factor C o fen6meno de megaprojectos. Atraidos pelosbeneficios fiscais e pelos recursos naturais, varios grandes investidores vieram a Moqambique depois de 1997, comeqando com a fundiggo de aluminio da xxxi MEMORANDO ECON~MICO DO PAiS MOZAL. Em2002, os megeprojectos representavam 7 porcento do PIB. A sua contribuiqiio anual ao crescimento do PIB ascendeu a 1,6 porcento. Espera-seque represente 10-11porcento do PIB ate ao fim da dtcada. Os megaprojectos ajudaram "a colocar Moqambique no mapa" e provavelmente tambem estimulargo investimento fora de megapvojectos. Todas estas realizaqdes foram positivas em grande ou menor medida e nenhuns efeitos negativos foram observados. Dai que se pode dizer que ha, pel0 menos, indicaqdesprimafacie de que apolitica de beneficios fiscais foi bem sucedida. xcix. Todos os factores descritos acima indicam que o crescimento foi sustentado por uma base alargada. Alem da agricultura, algumas partes do sector formal privado tambem registaram um crescimento substancial -manufactura, transportes, comunicaqdes, serviqos, extracqiio mineira -onde as taxas medias de crescimento (em termos reais) durante o period0 1995-2003 tiveram uma variaqzo anual entre 5 porcento e 15 porcento. Emparte devido ao restabelecimento dos mercados tradicionais da regiiio tais como turismo, hotelaria e restauraqgo. Alguns constituir- novos investimentos, a maior parte dos quais dos paises vizinhos. Este foi um crescimento sustentado por uma base menos alargada e, portanto, estas taxas s20 provaveis de nZo registar nenhuma alta significativa no futuro. 0 numero de pessoas activas no sector informal urbano registou umrapido crescimento a umataxa de 7-8 porcento, impulsionado pelamigraqso rural- urbana e um rapido crescimento da populaqiio em idade activa. C. Como e` que o indice depobreza bakou? Usando a linha de pobreza do Governo, o indice de pobrezabaixou dramaticamente, de 69 porcento em 1996/7 para 54 porcento em2002/3. A queda foi mais rapidanas zonas rurais (de 71para55 porcento) do que nas zonas urbanas (de 62 para 52 porcento). A pobreza continua sendo na esstncia um fenomeno rural, pois, a maior parte da populaqiio e rural. ci. A pobrezatambim pode ser avaliada com basenos bens, acesso aos serviqos e nos resultados sociais. Nestas areas tambem registaram-se melhorias. As taxas de propriedade de aparelhos de radio e de bicicletas subiram. 0 acesso a agua salubre aumentou de 24 porcento para 37 porcento. Em 1996/7, 8 porcento de todos os adultos tinham concluido o ensino primario do segundo ciclo (EP2); tendo aumentado para 11 porcento em 2002/3. 0 indicador-chave de saude e a taxa da mortalidade infantil, que baixou de 149 em 1995 para 101 em 2003, um dos mais rapidos decliniosjamais observados em Africa. cii. De ummodo geral, o consumo, determinado com base em pesquisasjuntos das familias, aumentou a uma taxa media anual de 4,6 porcento. AlCm disso, o consumo aumentou em todo o espectro de rendimentos. Por exemplo, o consumo dos 20 porcento da camada de populaqiio mais pobre aumentou em 23 porcento entre 1996/7 e 2002/3, ou seja, umcrescimento anual de 3,5 porcento. ciii. Quanto 21s regides, o indice de pobreza diminuiuem todas as provincias com a excepqiio de Cab0 Delgado e da provincia de Maputo. Estas constataqdes anomalas podem, contudo, ser descartadas, sendo o cas0 de Cab0 Delgado devido a uma amostragem deficiente durante apesquisa familiar; enquanto que na provinciade Maputo foi devido ao facto da comida consumida na area ser mais cara, sendo de uma qualidade xxxii SUMARIO EXECUTNO mais alta." Se fosse possivel normalizar a qualidade alimentar, a taxa de pobreza na provincia de Maputo poderiamuito bemter baixado durante o periodo compreendido entre 1996/7 e 200213. civ. Varios factores de politica relevantes podem ser identificados como tendo contribuido consideravelmente para a reduqBo da pobreza entre 199617 e 2002/3: k Doponto devista sectorial, o factor dominantenaexplicaqgo do desempenho do pais na redu@o da pobreza 6 o bom desempenho das familias camponesas. Isto por causa da vasta extensgo do sector agricola em relaqBo a outros sectores de emprego. P 0 aumento das taxas no sector de educaqZio contribuiupara areduq3o da pobreza. Isto porque os niveis da educaqgo aumentaram, e porque o impacto da educaqgo nos salirios (e no consumo) e consideravel. 0 impacto de umano adicional de escolaridade sobre o salario estanaordem de 3 porcento na agricultura nas zonas rurais e entre 5 porcento e 15 porcento em actividades nBo agricolas nas zonas mrais, e tambkm nas zonas urbanas; e o impacto aumenta ao nivel de ES2 e superior. cv. RelagTo entre o crescimento e a reduqko dapobreza. Como e que o processo de crescimento contribuiu para a reduqgo da pobreza no periodo entre 1996 e 2003? Primeiro, e mais importante, deve-se observar que enquanto o PIB cresceu em 8 porcento no periodo 1992-2003, o consumo privado determinado pelas contas nacionais cresceu a uma taxa mais modesta de 2-4 porcento, dependendo do periodo de tempo especifico.16 (Sendo a disparidade devida ao aumento das taxas de investimento e ao aumento acentuado das exportaqdes associadas com os megaprojectos.) Foi o crescimento deste consumo que levou a reduqgo que se observa no indice de pobreza. cvi. Emsegundo lugar, importa observar que o crescimento foi "pro-pobres". Paraos propositos do Memorando Economico do Pais, o crescimento que favoreqa as camadas pobres e definido como sendo a taxa media de crescimento do consumo das pessoas abaixo da linha de pobreza. Esta "taxa de crescimento pr6-pobre" 6 fortemente positiva (3,9 porcento por ano). Tambem n3o devemos esquecer que nenhuma regia0 ou ocupaq2o importante foi ignorada no processo de crescimento. cvii. A relaqgo especificaentre o crescimento e a reduqBo da pobreza podem agora ser enumeradas. Conforme indicado acima, o crescimento da produqgo agricola, devido em parte a expansgo da area e emparte B melhoria da produtividade, teve um impacto directo na reduqgo da pobrezanas zonas rurais (e em algumas zonas urbanas). Os empreendimentos nBo agricolas nas zonas mrais tambem tiveram um impacto directo na reduqBo da pobreza, em particular, nas tr&sprovincias mais meridionais com ligaqdes mais estreitas com o mercado. 0 aumento no consumo observado a todos os niveis e suportado pel0 aumento dos rendimentos. Os rendimentos de emprego17aumentaram consideravelmente entre 1996/7 e 2002/3: os rendimentos medios de emprego da familia l5Vide o texto principalpara mais analise destes dados. l6 cifrasSZOinferioresemrelaggoaosresultadosdolevantamentodoagregadofamiliar de 199617e Estas 2002/3. Isto nHo implica necessariamente que qualquer delas esteja errada. Vide a discussiio na nota de rodape 45, pagina 14. l7"Os rendimentos deemprego" aquireferem-se ao somatorio dosrendimentos de emprego nossectores formal e informal. A propor$io exacta das duas n5o e conhecida. Vide a nota de rodapt 4. xxxiii MEMORANDOECON~MICOP A k DO foram de Mt 5.500/dia, em 1996, aumentando para M t 15.000 em 2002/3.'* cviii. A medida que a economia crescia, as mudanqas estruturais ocorriam de acordo com as previsdes. 0 valor acrescentadoregistouum crescimento mais rapido na area de serviqos e, consequentemente, o crescimento do emprego foi mais rapido nesta area. 0 valor acrescentado registou umcrescimento mais lento na agricultura e, consequentemente, o crescimento do emprego na agricultura foi mais reduzido e as pessoastransferiram-se das zonas rurais para as urbanas. cix. Durante este periodo, os gastos do Govemo - com o apoio dos doadores - concentraram-se nas prioridades do PARPA, ou seja, educaqzo, saude, vias de acesso, agua e agricultura. Cada uma destas areas de despesas, directa ou indirectamente, contribuiu para a reduqgo da pobreza: 9 0 aumento dos gastos naeducaq8oelevaramataxa (bruta) das matriculasno ensino primario do primeiro ciclo de 56 porcento em 1995 para 110 porcento em 2003. A taxa de concluszo no EP1 aumentou de quase 20 porcento em 1990 para 40 porcento em 2003. 0 n6mero de adultos que concluiram o ensino primario aumentou de 8 porcento em 1996/7 para 11 porcento em 2002/3. 0 aumento nos niveis de escolaridade melhorou o nivel de rendimentosno mercado de trabalho e elevou a produtividade dos operadores do sector informal, atraindo maiores rendimentos nos empreendimentos rurais n%oagricolas. h Os gastosnasa6de aumentaram o acesso as instalaq6es e cuidados de saude (por exemplo, maior cobertura de vacinaqdes) e, assim, contribuiu directamentepara a reduqgo da pobreza. AlCm disso, a situaqzo de saude melhorada das pessoaslevoua melhorias na sua produtividade, o que tera presumivelmente contribuido indirectamente para a reduqgo dapobreza, pois, os adultos activos tCm agora maiores probabilidades de ter acesso a assistencia medica quando necessario e a taxa total de fertilidade registou umdeclinio em consequCnciada queda na taxa de mortalidade infantil libertando, assim, mais mulherespara o mercado de trabalho. No entanto, os esforqos com vista a conter a epidemia do HIV/SIDA ainda n5o tiveram sucesso (vide o paragrafo cxi, pagina xxxv abaixo). 9 Comos gastosnareabilitaqgo e manutenqzo de estradas, entre 1992e 2003, as vias que se encontram agora emboas ou razoaveis condiqdes aumentaram de 10porcento para 70 porcento da rede classificada e o numero de vias intransitaveis reduziu de 50 porcento para apenas 5 porcento. Tendo, assim, umimpacto directo napobreza, atravCs da geraqgo de emprego na reabilitaqgo e manutenggo das vias e tambem indirecto, atraves do acesso a transporte mais barato, margens mais baixas, maior comercio, melhores preqos para os produtores agricolas e preqos de bens mais baixos para o consumidor nas zonas reconditas. 9 Os gastosnos recursos hidricos aumentaram o acesso a aguadeumafonte melhorada de 12 porcento para 27 porcento nas zonas rurais e de 54 l8 cifrass80constantes2002Mt.Problemasdedefiniqzoafectaramadeterminaqlodosrendimentos Estas do sector informal e, dai, tambCm os rendimentos totais. Por isso, aqui enfatizamos: (a) o consumo, e (b) o rendimento de emprego. xxxiv SUMARlO EXECUTIVO porcento para 64 porcento nas zonas urbanas, entre 1996/7e 2002/3.19Isto tera provavelmente reduzido o indice de doenqas de transmissgo hidrica(por exemplo, diarreia e colera). cx. Emduas Areas, arelaqgo entre o crescimento e areduggo dapobrezango foi t3o forte como se esperava: > 0 sector formal geroupoucas oportunidades deemprego. A elasticidade do crescimento do emprego no sector formal foi de 0,l ,20 uma cifra baixa quando comparada com outros paises. Umfactor subjacente a este fenomeno podera ser a legislaqgo laboral restritiva que faz com que a expansgo do emprego ngo seja uma opqiio atractiva. (Vide comentarios adicionais sobre a lei do trabalho na pagina 40). > Os megaprojectos contribuiram significativamente na adiqiio de valor, mas ate ao momento t2m tido pouco (embora crescente) impacto na geraqgo de postos de emprego ou nas actividades de negocio locais via ligaqbes inversas. Calcula-se que em 2010, os megaprojectos responder30 por cerca de 2 porcento do emprego no sector privado e espera-se que no futuro as 1igaqBes inversas venham aumentar. A principal contribuiqiio dos megaprojectos na reduqgo da pobreza e atraves das suas contribuiqdes fiscais, que se espera aumentem 3 ou 4 porcento das receitas do Governo em 2010. Isto reforqa ainda mais a ideia de que as despesas do Governo t6m que ser feitas da forma mais eficiente possivel de modo a fazer us0 apropriado destas receitas incrementadas. cxi. Finalmente, ha uma area de resultados relevantes a pobreza em que nem o crescimento nem esforqos do Governo puderam (ainda) abrandar a sua deterioraqgo: HIV/SIDA. A prevalhcia da doenqa aumentoude 12porcento em 2002 (Amdt, 2003) para 16.2 porcento em 2004. 0 impacto da epidemia provavelmente vai se tomar mais serio naproxima dkcada. Em 2010, a esperanqa de vida ao nascimento baixara de 43 para 36 anos, ao invCs de aumentar para 50 anos. Os custos macroecon6micos e sociais s50 enormes. A abordagem de modelaqao indica que a propagaqgo da SIDA reduz anualmente o crescimento per capita do PIB em 1porcento, por causa do crescimento lento da produtividade, reduzida acumulaqgo de capital humano e reduzida acumulaqiio fisica, para n8o falar do dramatic0 fardo social da doenqa. cxii. 0 Governo est&aenvidar esforqosparadar respostaaeste desafio. Jafoi desenvolvida uma estratkgia nacional de HIV/SIDA e foi estabelecido um secretariado nacional de HIVISIDA. Os planos de mitigaqiio desta doenqa estgo a ser implementados nos principais ministerios. Estas acqbes t6m merecido todo o apoio dos doadores. Todavia, o nivel de implementaqgo tem sido baixo - tendo a execuq8o do plan0 sido implementada em apenas 70 porcento em 2004 e apenas USD 5 milhbes dos fundos do project0 de HIV/SIDA financiado pel0 Banco Mundial foram gastos nos dois primeiros anos. Hauma necessidade urgente de se melhorar a formaq5o e a gestgo. ''Ascifias s80 baseadas nos levantamentos do agregado familiar. Vide a discussiio no paragrafo 270 (p. 80) e nota de rodape 128 (p. 80) sobre o acesso a Bgua de fontes melhoradas. 2o Este refere-se a empresas n5o afectadas pela estruturap8o e privatizaG5o das empresas estatais e m meados da decada de noventa. Vide a discussiio no paragrafo 120 (p. 40), e a nota de rodape 85, e Benito- Spinetto and M o l l (2005) para mais pormenores. xxxv MEMORANDO ECONOMICO PA~SDO Desafios para o crescimento sustentado e para a reduqiio da pobreza cxiii. MoGambiqueparece estar bem posicionado para experimentar umperiod0 longo de crescimento per capita sustentado e tambem para alcanqar algumas dos principais Objectivos de Desenvolvimento do MilCnio, como a reduqzo do indice de pobreza em metade ate 2015. Todavia, os desafios com que o pais se confi-onta SZOinumeros. Existe, no entanto, a possibilidade de com umprogresso acelerado em certas reformas-chave, a taxa de crescimento poder atingir niveis elevados de entre 7 e 8 porcento a longo prazo, acelerando assim a reduqZo da pobreza. Os provaveis ceniirios de crescimento futuro - e os cenirios correspondentes de pobreza- si30 discutidos abaixo (xxvi); porem, antes de se estimar a taxa global de crescimento importa analisar os factores que influenciarzo os modelos de crescimento no futuro e considerar o papel do Govemo na abordagem das ameaGas a um crescimento sustentado, bem como o seu papel no aproveitamento de novas oportunidades de crescimento. As secq6es que se seguem exploram estes aspectos. cxiv. Os desafios que enfrentam os dirigentes moqambicanos podem ser divididos em tres categorias: gestzo macvoecondmica, desenvolvimento do sector privado e recuvsos naturais. Desafios ao crescimento e reduqlo da pobreza: gestlo macroeconbmica cxv. Foi mencionado anteriormente que o bom desempenho macroeconomico de Moqambique constitui um dos pilares dos seus indices de crescimento extraordinarios durante a ultima decada. No entanto, ainda prevalecem riscos em duas Areas relacionadas: estabilidade macroeconomica e gestzo das finanqas publicas. Sem melhorias nestas areas, as elevadas taxas de crescimento do pais sera0 comprometidas. cxvi. Estabilidade macroeconbmica. Actualmente, e do conhecimento geral que a estabilidade e previsibilidade macroecon6micas s50 o principal sustentAculo da reduqzo de pobreza e do crescimento, compreendendo a estabilidade dospreCos e o control0 fiscal. Os mais desfavorecidos s30 os que sentem o impact0 imediato da subida da inflaqgo, pois, esta actua como um imposto regressivo. A m6dio prazo, sZo tambem os pobres os mais afectados, pois, os investidores SZOafugentados e as oportunidades de crescimento e de geraqzo de postos de emprego entram em declinio. cxvii. NZo obstante o crescimento e progresso do pais na fiente macroeconomica, existe ainda o perigo da instabilidade dos preqos. A reorganizaqzo do sistema bancario em meados da decada de noventa, juntamente com o aperto monetArio, resultou na inflaqzo de um unico digito ate 1999. Porem, houve duas fases de inflaqgo provocada pela politica monetaria branda, uma com inicio em 2000 que resultou numa inflaqZo de 17 porcento em 2002, e outra com inicio em 2003/4 que resultounuma inflaqZo na ordem de 17 porcento no principio de 2004, baixando posteriormentepara 8 porcento em Janeiro de 2005. Nestes casos, a politica monetariateve de ser restringida atravCs de umaumento nas necessidades de reservas, aumentando a taxa banchria e/ou acelerando a venda das reservas em divisas ao p6blico. cxviii. Comparado com os padrdes dos paises menos desenvolvidos, Moqambique manteve uma boa disciplina fiscal na medida em que somente em alguns casos recorreu a emprestimos do sistema banchrio para compensar o defice fiscal.210 stock da divida 21 Umaexcepqii notavel a este desempenho positivo ocorreuem2001-2 quando dois bancostiveramque ser socorridos. 0 pais esta ainda a suportar os custos de "diferimentos ineficientes" que levaram a crise xxxvi SUMARIO EXECUTIVO intema phblica 6 de cerca de 4 porcento do PIB. cxix. 0problemaactualdo ladofiscal equeo desempenhodasreceitasabrandou durante 2004 e necessita de melhorar porque as necessidadesde gastos do Govemo na educaqgo, saude e infra-estrutura provavelmente continuariio a aumentar, pel0 menos, a par e passo com a populaqzo. AlCm disso, existe aprobabilidade da ajuda como uma percentagem do PIB vir a baixar do seu actual nivel elevado de 15 porcento. 0 plan0 do PARPAvisava awnentar as receitas para 15,4 porcento do PIB em 2005; em2004 tinha conseguido somente 12porcento. Dai o imperativo de se aumentar as receitas intemas do pais de modo a que o pais possa apoiar os seus programas de reduqiio da pobreza e, enfim,reduzir a dependencia de ajuda. cxx. A divida extemaphblica foi consideravelmente reduzidaatraves de duas operaqaesHIPC. Associado As excelentes perspectivas do pais em termos de exportaqaes, preve-se que o valor actual liquid0 do serviqo da divida estrangeira dividido pelas exportaqdes venha a baixar de 84 porcento registado em 2004 para 50 porcento em 2020. Os valores de 125-150porcento SZOconsiderados no context0 do HIPC para denotar umproblema de dividapendente. Dai que de acordo com este padrgo, Moqambique esta livre de divida extemapendenteque possa abrandar o crescimento. cxxi. 0 sistema cambial deMoqambique assentanumaflutuaqzo dirigida. A taxa de cambio efectiva depreciou ligeiramente desde a decada de noventa. Embora os pagamentos no gmbito da ajuda financeira tenham totalizado 12-19 porcento do PIB, isso n5o causou nenhuma apreciaqzo significativa da moeda. Ummodelo de equilibrio geral desenvolvido para testar isso indica que o aumento naprodutividade gerado pelos fundos da ajuda (por exemplo, estradas, educaqzo) compensou a pressgo para o aumento de preqos de bens n5o comercializhveis ("non-tradables"), evitando desse modo que o pais seja exposto ao efeito pemicioso da "dutch disease" nas exportaqdes. 0 ponto central e que a taxa de cimbio deve continuar a ser receptiva Bs tendCncias do mercado. cxxii. Gesta`odasJinanqaspublicas. A responsabilidade fiduciaria de Moqambique pode na melhor das hipoteses ser descrito como debil, embora o Govemo esteja a tomar medidas com vista a melhorh-la. As Areas mais fracas szo a contabilidade, auditoria e contratos publicos. Esta 6 por si so uma questgo de preocupaq8o. Tambem sera vital para o crescimento a proxima dkcada, porque: > a melhoria da eficiCncia nos gastos e necessaria para efectivar uma melhor provisiio de serviqos que, por sua vez, e necessaria para impulsionar o crescimento, em particular, a infra-estrutura, a educaq8o e a saude. Constatou-se que o custo das obras publicas C significativamentemais elevado em Moqambique (20-30 porcento) do que nos restantes paises da r egiiio; > as despesas totais nZo podem mais expandir ao ritmo anterior. E provavel que as despesas globais se situem em 24-25 porcento do PIB. Por isso, havera necessidade de, no futuro, se introduzir melhorias na eficisncia; 3 o financiamento dos doadores contribuiu parao crescimento e para a redu@o da pobreza e sem melhorias constantes na gestiio das finanqas bancaria, pois, as obrigaGdes n8o foram readquiridas e a emiss8o de novas obrigaqdes C umfactor contribuinte as elevadas taxas de juros. xxxvii MEMORANDO ECONOMICO DO PAiS publicas partes dos fundos dos doadores poderaestar emrisco. Isto e especialmente importante porque Moqambique recebe mais ajuda do que muitos outros paises. A justificaqgo para tal generosidade consiste no facto da ajuda estar a ser bem usada- conforme evidenciado pelas elevadas taxas de crescimento e melhorias significativas nos indicadores sociais. No entanto, se esta situaqgo se alterar negativamente, ajustificaqgo especial de Moqambique vai deixar de ser e o seu cadastro fiduciario bastante debil pode bem resultar na reduqgo da ajuda para os niveis normais. cxxiii. 0 Govemo empreendeuumprograma sequenciado de reformas, que comeqa com uma nova lei de gestgo financeira em 2001 e regulamentos acompanhantes em 2002, que estabelecem a base tanto para os procedimentos contabilisticos modemos como para a reforma dos contratos pdblicos. 0 sistema de gestao financeiro electronic0 SISTAFE esta a ser implementada numa fase piloto em dois ministerios em 2005. Depois sera alargado aos outros ministkrios sectoriais em 2006. Enttio seguir-se-a o melhoramento dos classificadores (programa, funcional, regional) de modo a fomecer os detalhes suficientes para permitir a analise dos impactos da pobreza. Concomitantemente, os regulamentos e as instituiqbes que regemos contratos pGblicos est50 a ser revistos com vista apromover a concorrencia e a transparencia. cxxiv. Uma outra tarefa importante t a coordenaq5o da implementaqzo do SISTAFE com a reforma do sector publico e, em especial, pennitir a concord?" entre o quadro de funcionarios mantido pel0 MinistCrio da Administraqgo Estatal e a folha de pagamentos do MinistCrio de Finanqas. Assim sera possivel eliminar os pagamentos a funcionarios que efectivamente ngo existem: o exame das despesas publicas do Govemo no sector de educaqzo em 2003 concluiu que no ensino primtrio ntio houve uma prestaqgo de contas adequada de cerca de 20 porcento dos funcionarios.22 Desafios ao crescimento e reduggo da pobreza: desenvolvimento do sector privado cxxv. Aqui referimo-nos a agendahamuito pendente de reformas visando apromoqgo da iniciativa do sector privado na actividade empresarial e remoqgo das barreiras ao investimento de firmas nacionais e estrangeiras. Poucamateria nesta secqgo e distintamente nova. A contribuiqgo do Memorando Economico do Pais na area de desenvolvimento do sector privado tem sido a concepqgo apriorizaqiio viavel de tarefas - legislativa, regulatoria, administrativa -para facilitar o investimento. Estapriorizaq5o encontra-se de forma resumida concisa napagina 68). Nos paragrafos que se seguem, s80 discutidas algumas das principais intervenqbes: P facilitaqgo do comercio P reforma e investimento no sector de energia P acesso financeiro. cxxvi. Facilitapio do comkcio. A rapida expansgo das exportaqbes foi uma importante fonte do crescimento durante os ultimos 10 anos em Moqambique. Desde o principio da decada de noventa, as exportaqbes est50 a expandir a u n a taxa (10 porcento) substancialmente mais alto do que as exportaqbes mundiais (6 porcento). Moqambique e umdos poucos paises emAfrica que conseguiu aumentar a sua fatia das exportaqbes mundiais. Estarapida expansgo das exportaqbes foi conduzida pela iniciativaprivada, 22 Vide a discussgo sobre o emprego no sector da educa@o em World Bank (2003b). xxxviii SUMARIO EXECUTIVO tanto nacional como estrangeira. cxxvii. A semelhanqa de outros paises, para Moqambique, uma estrategia de crescimento conduzida pel0 sector privado baseadanas exportaqdes tem o maior potencial a longo prazo para sustentar uma elevada taxa de crescimento a luz das condiqdes existentes: o mercado interno 6 limitado e detem umpoder de compra relativamente baixo. Alem disso, dado que a taxa das poupanqas nacionais ngo e suficientemente alta para gerar um crescimento adequado para ter umimpacto significativo napobreza, deve-se colocar Cnfase na atracqgo do investimento directo estrangeiro. cxxviii. Entre as razdes do sucesso de Moqambiquerelativamente as exportaqdes e a persistencia na reforma. Moqambique liberalizou significativamente as suas importaqdes. A mediaponderada comercial dos direitos de importaqgo e de 9 porcento. 0 Govern0 pretende reduzir a taxa maximapara 20 porcento em 2005. Moqambique esta empenhado numanova liberalizaqilo no Bmbito do Protocolo sobre Trocas Comerciais da SADC e pretende negociar umanovareduqZo de tarifas no context0 do Acordo de Parceria Economica (APE) com a Unigo Europeia (UE). cxxix. Muitos desafios prevalecem. Foi observado acima que a maior parte do crescimento das exportaqdes foi devido aos megaprojectos e que as exportaqdes "tradicionais" cresceram a ritmo lento. Embora a produqgo de aluminio, gas, titBnio e electricidade continuara a ser a principal fonte de crescimento das exportaqdes, o principaldesafio ao alargamento da base de crescimento do valor acrescentado para alcanqar os objectivos do alivio da pobreza C a remoqgo das barreiras para permitir uma maior contribuiqilo das pequenas e medias empresas. cxxx. Muito ha ainda por se fazer com vista a melhorar os serviqos de apoio as exportaqdes de modo a eliminar o vies anti-exportaqgo inerente ao sistema dos direitos de importaqgo. As prioridades Go: extensgo dos incentivos a exportaqao As pequenas empresas, a aceleraqgo do reembolso do impost0 sobre o valor acrescentado (IVA), melhoramento do acesso das empresas ao sistema financeiro e melhoramento dos padraes e da capacidade de satisfaqgo das medidas sanitarias e fitossanitarias, em particular, do mercado da UE. Considerando que as condiqdes macroeconomicas e fiscais est80 a melhorar e o regime das taxas de ciimbio 6 de ummodo geral adequado, os principais constrangimentos relativos a promoqgo de um crescimento conduzido pelas exportaqdes tCm a ver com as "questdes intemas": > o sistema de infra-estrutura dCbil (incluindo electricidade e estradas), que ainda necessita de investimentos avultados e melhoramento das politicas, e P o dificil clima de investimento (por exemplo, o desembaraqo lento das importaqdes, procedimentos de tributaqgo complexos, process0 lento de arranque dos negocios, um sistemajudicial que nZo funciona, custos elevados de transporte e alto risco de emprestimos). cxxxi. Reforma e investimentono sector de energia. Uma area requerendo uma forte orientaqgo politica e investimento substancial e a de energia. Cerca de 64 porcento das empresas da industria de manufactura acham que a electricidade representa o mais sCrio problema em termos de infra-estrutura. Reportam haver cerca de 17,5 interrupqdes mensais no fornecimento de energia. 0 acesso a electricidade aumentou de 4 porcento xxxix MEMORANDO ECONOMICO PAIS DO em 199617 para 7 porcento 23em200213 (240.000 familias), mas a maior parte das novas ligagaes ocorreram no quintil das familias mais abstadas, cujo acesso aumentou de 11 porcento para 18 porcento. Emtodo o caso, o ritmo de novas ligagaes -no ritmo historic0 de 8.000 por an024- e s6 ligeiramentemais alto do que o da populaggo, pel0 que a baixa taxa de electrificaqgo podera vir a estagnar. No context0 actual, 80 porcento das familias mais pobres ngo tergo acesso a electricidade por muitas dCcadas. Esta fitura estagnaqgo no acesso A electricidade contrasta negativamente com as melhorias previstas nas areas das telecomunicaq6es, agua, estradas e outros aspectos relacionados com infra- estrutura. cxxxii. E necessaria uma abordagem integrada e coordenada, combinada com medidas de politica, reestruturaggo e aumento de investimento. A curto prazo, P Deve-se melhorar o desempenho financeiro e operacional da Electricidade de Moqambique (EDM) e reduzir as perdas de energia de 20 porcento para 8-10 porcento. P Os serviqos ao consumidor devem ser bastante melhorados, por exemplo, atravks de umamaior expansgo dos serviqos prC-pagos, melhorando dessa forma a colecta de receitas. > Deve-se aumentar o investimento de modo a intensificar as ligaqaes onde a espinha dorsal das linhas de abastecimento ja exista. A rede deve ser reabilitada e reforgada. A medio prazo, P As reformas institucionais e legais devemvisar o estabelecimento de um ambiente favoravel conducente ao desenvolvimento do sector, com capacidade regulatoria adequada. cxxxiii. Melhovia do acesso e do custo definanciamento. A grande maioria das empresas em Moqambique indica que a falta de acesso ao financiamento e o seu custo elevado constituem um grande problema ao seu crescimento. As taxas dejuros si50 mais elevados quando comparadas com paises no mesmo nivel de desenvolvimento. Isto e em parte porque os "spreads" das taxas dejuros s80 elevados (19 pontos percentuais), e por causa de elevadas perdas dos emprestimos e dos elevados custos administrativos. So 29 porcento das empresas tCm acesso a emprestimos bancarios, inferior aos niveis em outros paises no mesmo patamar de desinvolvimento (por exemplo, a Eritrea com 45 porcento das empresas). As pequenas empresas enfrentam barreiras ainda maiores. Enos poucos casos em que as pequenas empresas superam as barreiras, elas enfrentam exigsncias colaterais estritas, a uma media de 141porcento do montante de credito. cxxxiv. Por isso, si50 indicadas varias reformas. A supenisgo bancaria deve ser reforqada ainda mais, harmonizando a classificaqgo dos emprkstimos com a pratica intemacional e desenvolvendo as competCncias centrais dos funcionhrios responshveis pela supervisgo. E necessario estabelecer umtribunal comercial que lidar&com os maiores casos. Simultaneamente, maisjuizes devem ser formados na resoluqgo de diferendos comerciais e os procedimentos judiciais devem ser simplificados a fim de minimizar a complexidade 23 Vide Fox et al. (2005), Table 5. . 24 Vide de Barros (2005), parhgrafo52. Entre 1996 e 2000 (4 anos), havia 31.000 novas liga@es. x l SUMARIO EXECUTIVO processual. 0 Ambito e fiabilidade do registo de credit0 administrado pel0 Banco de Moqambique devem ser alargados. Para desenvolver o mercado de capitais, que e muito pequeno, o primeiro objectivo deve ser o de desenvolver ainda mais o mercado de titulos pcblicos e, desse modo, providenciar umponto de referenciabaseado no mercado para a emissiio de titulos privados. cxxxv. Finalmente, o melhoramento do acesso ao financiamento pelas microempresas tambCm contribuiria para a minimizaqiio do constrangimento impost0 pel0 elevado custo do capital em Moqambique. Embora a indhstria de microfinanqas tenha crescido, o seu alcance ainda e pequeno, com maior concentraqiio em Maputo. E necessario eliminar os nos de estrangulamento, incluindo as restriqbes tomada de depositos. A legislaqiio a actual napreparaqiio deve pennitir que instituiqbes de finanqas micro tomem depositos. Aconselha-se que se recorra aos serviqos especializados de peritos para a finalizaqiio da lei sobre micro-finanqas. Desafios ao crescimento e B reduqiio da pobreza: recursos naturais cxxxvi. Esta secq2o e de facto uma continuaqiio da anterior sobre eficiencia. Um elemento importante no melhoramento da eficiencia da economia napromoqgo de investimento e criaqiio de oportunidades para a geraqiio de rendimentos pelos pobres C o melhoramento do quadro regulatdrio dos recursos naturais, especificamente, emrelaqiio a agua, extracqiio mineira, terra rural, pescas e florestas. Estas areas siio priorizadas e agrupadas nesta secqiio porque (a) colectivamente, siio vitais para o crescimento economico, (b) s50 essenciais para a reduqBo da pobreza (especialmente, a agua) e para a subsistencia (florestas, pescas, terra rural), (c) o seu abuso resultaria na insustentabilidade ambiental e ameaqaria o process0 de crescimento no seu todo, e (d) alguns destes elementos servem de excelentes estudos de cas0 em termos de eficiencia, govemaqiio e perspectivas de investimento. cxxxvii. EmMoqambique, certas caracteristicas si50 comuns a estas cinco areas dos recursos naturais: 9 relut2nciaemusarprocedimentostransparentes paraa atribuiqiio dos direitos de us0 e aproveitamento dos recursos P colecta deficiente das rendas de aluguer decorrentes ao Estado, e, o consequente us0 ineficiente dos recursos, menos crescimento economico, receitas decorrentes ao govemo mais baixas e, consequentemente, menos reduqiio da pobreza. Assim, conforme sera explicado nos paragrafos que se seguem, as melhorias na gestiio dos recursos naturais siio essenciaispara a reduqiio da pobreza e para o crescimento. cxxxviii. Agua. De todos os recursos naturais, o que requer maior atenqiio dos decisores politicos e a agua: a irrigaqiio de pequena escala, a mitigaqiio das cheias e da seca e o abastecimento de agua salubre tem umpapel ajogar na reduqiio da pobreza. cxxxix. kgua de irrigaglo: Cerca de umdecimo dos 36 milhbesde hectares de terra aravel no pais e apropriado para irrigaqiio. A terra cultivada representa 4,3 milhaesde hectares, dos quais so 0,9 porcento siio efectivamente irrigados - a mais pequena proporqiio em toda a regi2o. cxl. No context0 de capacidade de annazenamento de agua limitada e de falta de recursos financeiros, o melhor us0 dos recursos piiblicos requereria a concentraq2o dos esforqos no melhoramento das condiqbes dos pequenos agricultores. Seria mais viavel xli MEMORANDO ECONOMICO PAiS DO concentrar as atenqbes nos sistemas de pequena escala, de baixo custo e nHo em grandes projectos. 0 foment0 da irrigaqzo de pequena escalatambem levaria a um aumento na produggo de culturas de elevado valor e o que, por sua vez, aumentaria os rendimentos das familias mais pobres. Embora exista umpotencial significativo para a expansso da irrigaqgo comercial, tais iniciativas devem ser promovidas pel0 sector privado. 0 papel do Govemo e desenvolver umquadro legal, institucional e regulatorio, bem como apoiar o desenvolvimento da infra-estrutura de armazenamento de Agua ligado a projectos de us0 multiplo (energia / agua/ transporte / gestHo de seca). cxli. Infra-estrutura de recursos hidricos. A capacidade uti1das barragens em Moqambique representa somente 5 porcento da media anual do escoamento das aguas dos rios do pais, coma excepqgo do Zambeze. Assim, o pais precisa de desenvolver a sua infra-estrutura para o armazenamento de uma quantidade maior do escoamento. cxlii. 0 Govemo pode fazer progressos significativosnareduqgo da vulnerabilidadea secas tratando das questaes relacionadas com o abastecimento da agua rural, de agua de irrigaqzo para os pequenos agricultores e para a operaqiio e manutenqgo dos sistemas existentes. Isto pode ser complementado por medidas visando melhorar a base de dados das informaqbes, incluindo a negociaqHo com os paises ribeirinhos a montante (para assegurar que a operaqzo de reservatorios a montante nZo contribua para o agravamento das inundaqaes no territorio moqambicano) e sistemas de aviso prCvio que serviriam tanto para a gestgo de secas como de cheias. Deve-se construir pequenas represas e reservathios em todo o pais nas areas mais propensas a seca. H a uma necessidade urgente de se construir duas grandes represaspara o abastecimento de agua a Maputo e Beira e para irrigaqgo no Baixo PunguC. SHo tambCm necessariasrepresas de media dimensgo para assegurar o abastecimento de agua a varias cidades pequenas. Para a atenuaqzo de cheias, e necessario construir reservatorios de armazenamento, condicionada a reviszo das normas de operaqgo de todas as novas e antigas represas e ajusta-las de forma a equilibrar as necessidadesemtermos de atenuaqiio das cheias e outros usos. cxliii. Abastecimento de rigua. A falta de agua salubre e um factor nocivo a boa saude e a fraca produtividade do trabalho. Aqui exploramos o trabalho realizado no 2mbito da Reviszo das DespesasPublicas de 2003 - informados pela dramatica constatapo de que o nivel de acesso A agua salubre na zona rural era mais baixo do que se tinha sido pensado (27 porcento e ngo 40 porcento) e que o de acesso a agua salubre na zona urbana era mais alto do que se tinha sido pensado (64 porcento e niio 38 p ~ r c e n t o )Ha . ~ ~ necessidade de muitas melhorias, em particular, nas zonas rurais. cxliv. As iniciativas recentes no abastecimento de hguaurbanaprovaram ser um sucesso. Desde 2002, a cobertura das ligaqaes domesticas nas zonas urbanas era somente de 25 porcento e a agua estava disponivel so 11horas por dia. 0 govemo optou por um modelo de "gestgo delegada", introduzido atraves de um concurso publico em cinco municipios. 0 sector privado iniciou as operaqbes em 2002. Numperiod0 de dois anos tomou-se evidente que a decisiio tomada tinha sido correcta: a cobertura melhorou de 25 para 30 porcento e a provisHo aumentoupara 15 horas e o sector da agua urbana atraiu 25 Cifras baseadas nos levantamentos do agragado familiar. 0 acesso a agua de fontes melhoradas nas zonas rurais aumentou de 12 porcento em 1996i7 para 27 porcento em 200213 e nas zonas urbanas de 56 porcento em 199617para 64 porcento e m 200213. Vide a discuss50 no paragrafo 270 (p. 80) e a nota de rodape 128 (p, 80) sobre o acesso a Bgua de fontes melhoradas. xlii SmAruoEXECUTIVO novos investimentos. cxlv. Concomitantemente, o govern0 adoptou a "abordagem orientadapelaprocura" nas zonas mrais, que requeria que as comunidades fizessem uma solicitaqgo especifica para a instalaqgo de uma fonte de agua, pagando uma taxa e assumindo a responsabilidade pela sua manutenqzo. Mesmo assim, normalmente cerca de 30 porcento dos poqos deixam de funcionar por falta de peqas sobressalentes. cxlvi. Apesar do progress0 evidente registado, continuar com o actual mecanismo 6 insuficiente. Embora os niveis de acesso educaqgo e saude basicas em Moqambique seja igual ou superior a media afiicana, o nivel de acesso a agua de uma fonte melhorada nas zonas mrais e de 27 porcento, inferior a media africana de 46.5 porcento. cxlvii. Uma estrategia integradade rhpida expansco do acesso a agua salubre deve ter em conta dois principios: custo-eficbcia e a "abordagem orientada pelaprocura" -pois, sem establtima, aproporqi3o de poqos ngo operacionais rapidamente ultrapassara os actuais 30 porcento. Alem disso, devem ser concebidos mecanismos para assegurar a disponibilidade de peqas sobressalentes e a capacitaqgo da comunidade. Esta gigantesca expansgo pode ser alcanqada atraves de > Aumento de recursos e de pessoal no acesso a agua rural. Isto podera implicar a reafectaqgo de sectores menos prioritarios. Embora a abordagem continue a ser orientada pela procura, o principal constrangimento presentemente nZo SZOas comunidades que satisfazem os critkrios mas sim o pessoal, capacitaqgo e financiamento parapermitir a implementaqzo. > Investigar osfactores que s i 0 a base dos elevados custos de perfuraGZo. Emespecial, deve-se avaliar se a subcontrataqzo de mais trabalho atraves de concursos publicos, reduzindo assim o trabalho realizado pelas empresas estatais ngo reduziria os custos. > Expandir o acesso afontanarias comunais nas zonas urbanas. Isto pode ser feito a um custo mais baixo do que a reticulaqzo das casa. Umpossivel modelo de gestgo e a subcontrataqBoa particulares (como esta a ser feito em algumas cidades moqambicanas como Angoche). > Licenciar operadores privados de quiosques de agua como acontece em outras cidades africanas. k Aumentar a compartilha dos custosnus zonas urbanas. Hanecessidadede se aumentar as taxas em todas as cidades, em particular, nas cidades mais pequenas, a niveis que conseguem cobrir os custos operacionais e de manutenqtio e umaproporqzo do serviqo de divida. cxlviii. A extragh mineira est&a tomar-se umrecurso natural cada vez mais valioso. A recente aquisiqzo por uma firma mineira brasileira dos direitos de exploraqzo naregia0 carboniferade Moatize e uma indica950 das excelentes perspectivas que existem neste sector. Presentemente, o sector representa apenas menos de meio porcento do PIB e contribui receitas fiscais no valor de de $3-5 milhdes (alem do valor pago pela companhia de prospecqgo de cawgo). 0 numero de empresas duplicou desde2001. Isto 6 emparte devido a mudanqas legislativas progressivas e a introduqgo dum Cadastro de Exploraqgo Mineira inovador e transparente, que efectivamente constitui um exemplo das refonnas institucionais que e necessario introduzir (por exemplo, terra, pescas, florestas). A actividade mineira artesanal e de pequena escala proporciona uma ocupaqzo, em tempo xliii MEMORANDO ECONOMICOPAkDO integral ou parcial, a 100.000 pessoas espalhadaspor todo o pais. cxlix. Contudo, a produqzo mineira do pais encontra-se ainda muito abaixo da dos paises com potencial geologico semelhante. As despesas anuais de exploraqzo podiam ser aumentadas do seu nivel actual de US$15-25 paraUS$ 50-100 milhdes por ano. cl. Para conseguir isto, o sistema de tributaqgo mineira deve ser harmonizado com os planos fiscais do resto da economia. 0 sistema actual n%oe competitivo em termos globais. Ele consiste do Impost0 sobre o rendimento das pessoas colectivas - IRPC, mais os royalties e o imposto de licenqa e o Cddigodos Benefcios Fiscais. 0 sistema nzo e atractivo aos investidoresprivados porque os royalties podem constituir um grande encargo as empresas durante os periodos de declinio nos mercados de minerios e porque o IRPC n2o permite a acumulaqzo e os transporta das despesas durante as fases de desenvolvimento e de exploraqzo. 0 sistema de tributaqzo mineira deve ser revisto, a fim de torna-lo globalmente competitivo e assegurar umaproporqzojusta ao Governo. cli. Foram introduzidas na administraq2o da terra rural atravks de umanova lei (1997) e regulamento (1999) de terra: os mecanismos costumeiros de posse foram aprovados oficialmente e C necessario umprocesso de consulta antes da emiss2o de novas concessees. Desta forma a seguranqa de posse dos pequenos agricultores foi redobrada. clii. Estas reformas G o ainda insuficientes. Visto que a terra n8o pode ser adquirida, a unica barreira a obtenq2o dos direitos C meramente burocratica; n2o surpreende, portanto, que a concesszo de grandes areas de terra seja ainda feita a um custo praticamentenulo. A falta de ummercado oficial de terra cria oportunidades para as pessoasderivarem beneficios financeiros e C ineficiente porque os direitos a terra ngo gravitargo naturalmente aos seus utentes mais produtivos. Quando os melhoramentosnas zonas rurais s90 alienados, a concess2o de terra correspondente n20 6 transferida automaticamente com a venda.26 E necessario obter uma nova autorizaqzo antes que a concesszo possa ser transferida, o que cria novas oportunidades para as pessoastirarem beneficios financeiros do processo. Uma outra dificuldade tem a ver com o facto das concessdes estarem sujeitas a interferencia burocratica, devido a necessidade de umplan0 de us0 e aproveitamento de terra. Estainterferhcia cria uma percepqzo justificada da parte de potenciais agricultores de grande escala de que a posse de terra C demasiado insegura para merecer grandes investimentos. cliii. Podem ser introduzidas melhorias significativas na eficiencia atraves das medidas seguintes '9 reforma dos regulamentos parapermitir a transferencia automcitica dos direitos de us0 e aproveitamento de terrajuntamente com a venda de edificios e de melhoramentos, sem nenhumainterferhnciaburocratica. '9 aumentar substancialmente o imposto de terra para concessionarios de mCdia e grande dimens20, desse modo, obrigando os proprietiirios de terras que n5o estejam a fazer us0 produtivo da sua terra para transferi-la para outros que fariam. Os concessionarios podem ser compensados sendo aliviados da exigencia de planos de us0 e aproveitamento. Isto promoveria a percepqgo de seguranqa na posse de terra e estimularia mais investimentos. cliv. Aspescas representam 1,5 porcento do PIB e 10-15 porcento das exportaqdes; a 26 Veja nota de rodape 123 (pagina 74). xliv SUMARIO EXECUTIVO maior parte resultando da pesca industrial e semi-industrial de camarBo. A pesca artesanal e de subsistencia 6 tambem importante na medida em que proporciona emprego a cerca de 90.000 pessoas. Existemtrts grandes companhias industriais e cerca de 200 semi-industriais envolvidas na pesca de camargo. As taxas das licenqas ascendem a USD 5 milhdes anualmente, embora esta cifra esteja aqutm dos rendimentostotais devido a razdes que s50 elucidadas no corpo principal do texto abaixo. 0 Govemo tem estado a prosseguir uma estrategia de promoqBo de participaqBo local (dos nacionais) napesca de camarBo, atraves da Emopesca, empresa pertencente ao Estado, para formar empreendimentos conjuntos. Isto foi conseguido atraves da emissBo de um numero substancial das novas quotas para novos intervenientes sem, no entanto, reduzir a participaqgo da frota industrial. NBo surpreende, portanto, que todo o esforqo pesqueiro tenha aumentado por um factor de tres desde os finais da dCcada de setenta, embora o pescado total tenha permanecido quase ~onstante.~' clv. Todavia, o procedimento de atribuiqZo de novas licenqas nZo foi transparente, com valores substanciais a serem transferidos sem nenhumaregulamentaqgo Clara ou critkrios de distribuiqzo. Emtodo o caso, o mecanismo nBo teve o sucesso que se previa aquando da planificaqBo inicial, na medida em que os operadores moqambicanos nBo tCm conseguido desenvolver a sua quota e os operadores nacionais SZOAida muito poucos. chi. A perdade receitas decorrentes da exploraqBo deste recurso deve ser abordada. Isto pode ser conseguido atraves de ummecanismo dual. 0 acesso internacional podia ser melhor organizado atraves de um sistema de leilBo. Isto aumentaria de forma consideravel as receitas totais arrecadadaspel0 Estadoq2'A razBo da nBo adopqBo do metodo eficiente do "primeiro-melhor" do sistema de lei150 em todo o sistema C que este provavelmente excluiria quase completamente os operadores nacionais. Dai que para o acesso nacional, seja necessario um conjunto claro de criterios publicamente disponiveis de atribuiqgo de quotas, para assegurar a existtncia de sistemas previsiveisejustos para os utentes e tambkm assegurar os beneficios sociais da exploraqZo dos recursos naturais. Tal "apadrinhamento" deve ser eliminado por etapas dentro de umperiodo predeterminado de, por exemplo, cinco anos. clvii. A pesca artesanal e de subsistencia constitui uma importante zona tampgo para a pobreza. 0 desenvolvimento comunitario em geral foi iniciado como parte do desenvolvimento da pesca, com a assistencia do Instituto de Desenvolvimento Pesqueiro de Pequena Escala. A pesca artesanal esta sujeita a regulaqBo e ngo se propde que isso seja alterado. A pesca de subsistencia e de livre acesso, que e a melhor abordagem para este tip0 de actividade; C importante que zona tampBo nBo da pobreza nBo seja comprometida pela gestgo inadequada ou outras intervenqdes govemamentais. NBo existe nenhumaindicaqBo de que o livre acesso neste cas0 leve ao depauperamento do recurso. clviii. Asflorestus representam 2,5 porcento do PIB. A produqgo deste subsetor e de cerca de 120,000 metros c6bicos por ano. Os peritos estimam o potencial em cerca de 500.000 metros cubicos. Foram introduzidas recentemente importantes mudanqas de politica. A taxa de denube de arvores foi aumentada significativamente, o que que fez 27 Uma vez que foi impost0 umperiodo de defeso, nZo risco de depauperamento aos niveis actuais de exploraqZo. 28 Vide o texto principal que contCm referhcia sobre os dados. xlv MEMORANDO ECONOMICO PAk DO com que as receitas totais aumentassem de US$ 0,5 milhdes em 2002 para US$2,3 milhdes em 2003. PrevC-se umnovo aumento de 100porcento da taxa em 2005. Os calculos plausiveis indicam que os niveis de taxas resultantes aproximar-se-Ho das taxas de escassez. U m a outra reforma significativa tem a ver com os esforqos das autoridades no sentido de se criar maior valor acrescentado localmente. Com o objectivo de estimular o processamento local de madeira - ao inves de exportaqHo de toros nHo processados - as autoridades impuseram, em 2003, uma proibiqHo na exportaqiio de madeira de alta qualidade nHo processada. clix. U m a desvantagem do actual sistema de taxas ("taxa de licenqa de volume") C que estabelece a diferenciaqHo com base no tipo de madeira, o que esta correcto, mas n8o pela localidade. Isto da incentiva indevidamente a exploraqBo de madeira proximo das zonas populosas e desencoraja o corte em zonas distantes. Este mecanismo 6, portanto, ineficiente, pois, causa a subutilizaqBo das areas remotas onde nZo existe nenhum problema ecologico e provoca o desflorestamento das areas de facil acesso. clx. A mCdio prazo, uma forma mais eficaz de estimular umamaior concorrCncia seria um regime de licenciamento com base em 1icitaqHo. Deste modo, ao inves do Govemo aplicar uma taxa unitaria baseada no toro, seriam os proprios madeireiros que (efectivamente) estabeleceriam os niveis das taxas e porque fa-lo-iam de forma competitiva, haveria um aumento substancial de receitas -possivelmente at6 13 milhaes de dolares. clxi. A proibiqgo da exportaqHo podia tambCm ser reexaminada. NZo esta claro que esta seja a melhor forma de realizar o objectivo de promoqHo do acrescimo de valor local. Por um lado, as instalaqdes fabris que actualmente existem nHo conseguem ainda processar toda a madeira disponivel e nHo C surpreendentemente, portanto, que a quantidade total de madeira produzida tenha baixado. Alem disso, ha uma perda de receitas potenciais, se a proibiqHo nHo for efectivamente policiada. Mais importante ainda, a relaqHo entre a proibiqHo da exportaqHo e os investidorespotenciais e debil e 6, portanto, improvavel que esta medida surta o efeito desejado. Seria mais eficaz e vantajoso em termos fiscais, se fosse introduzido um conjunto de medidas graduais e cautelosas em que se combinasse as vantagens de taxas fixas durante umperiodo de tempo, taxas de exportaqHo fixas durante umperiodo de tempo e brandas e inventivos em termos de infra-estrutura. clxii. 0 sector de florestas t tambem umimportante contribuinteno sustento de dezenas de milhares de camponeses. Por essa razHo, o actual sistema de emissHo de simples licenqas ao balcHo para o corte 500 metros cubicos de corte anuais deve ser mantido. Previsaes de crescimento e a pobreza clxiii. Qual e o desempenho do crescimento de Moqambique no futuro? Ummodelo de equilibrio geral sugere uma taxa media de crescimento a longo prazo de 7 porcento, da qual se deve subtrair 1porcento por an0 devido ao HIVISIDA e mais 1porcento por an0 para levar em conta o risco meteorologico, chegamos assim a uma taxa de crescimento a longo prazo de 5 porcento. Estataxa e melhor que a da regia0 situada em 3.4 porcento - e deve-se a superioridade de Moqambique em termos de recursos naturais. No entanto, esta taxa e inferior a do crescimento recente de 8 porcento verificado no pais. Seria arriscado, contudo, assumir uma taxa de crescimento "como de costume" de 8 porcento, pelas seguintes razdes: xlvi SUMARIO EXECUTIVO 9 tal crescimento aceleradonormalmentedura apenas alguns anos, > Moqambique enfrenta o risco meteoroldgico e outros riscos, 9 arecuperaq5o agricolaja atingiuo seunivelmaximo e o sector estaacrescer a umataxa mais baixa do que antes, 9 amaiorpartedasreformas"faceis" queestimulam o crescimento j iforam introduzidas, por exemplo, a liberalizaqgo das telecomunicaqbes. Isto deixa as reformas institucionais mais dificeis como a reformajudicial, e > a ajuda extema diminuira como umaproporq5o do PIB. clxiv. Com uma agenda de reforma vigorosa, por outro lado, pode ser possivel atingir taxas de crescimento de 7-8 porcento a medio e longo prazo. Podia-se iniciar um "ciclo virtuoso" a medida que as autoridades prosseguem com os melhoramentos institucionais com aquilo que pode ser denominado "reformas de segunda geraq5o". Para convencer os cepticos de que isto 6 possivel, citamos quatro exemplos de mudanqas institucionais de longo alcance - e as vezes dolorosas: 9 As reformasno subsector dos transportesadreos em2001levarama entrada de umnovo operador domestic0 e ao aumento de viagens turisticas. 9 A liberalizaqgodas telecomunicaq6es em2001 levaramaumgrande aumento: de 85.000 linhas fixas e 51.000 telemoveis, para a 700.000 telemoveis em 2004, com um melhor serviqo e os preqos mais baixos da regi5o. 9 A adopq5odeumCadastro deExploraqiioMineira transparente en5o discricionario, juntamente com umanova lei de minas favoravel ao investimento, resultounumaumento acentuado de requerimentos de titulos de explora@o mineira, de 110 em 2001 para 180 em 2004, e o tempo para a obtenq5o de licenqa reduziu de 120 dias em 1999para 50 dias em 2003. A base de sustentaqgo para o crescimento do sector a longo prazo ja foi estabelecida. 9 Confonneobservado acima, a``gestzo delegada" da aguaurbanalevoua melhorias importantes em termos de cobertura e periodo de abastecimento. Estas experihcias podem ser replicadas noutros sectores, ao longo das linhas sugeridas na secq5o dos "Desafios" acima. clxv. As simulaqaes feitas usando a literaturade "crescimento regressivo" sugerem que Moqambique pode melhorar o seu desempenho de crescimento melhorando o seu ambiente de neg6cios aos niveis da Asia Oriental. Podem ser conseguidos ganhos de at6 1ponto percentualpor ano evitando as crises bancarias, melhorando a govemaq8o ou reduzindo os gastos improdutivos do govemo. Ganhos maiores de ate 3 pontos percentuais por an0 podem ser assegurados - com o tempo - atraves do melhoramento da infi-a-estrutura ou aumentando o acesso ao ensino secundhrio. c h i . Caso o ritmo de crescimento do pais seja mantido e contanto que as baixar. Assumimos o crescimento do PIB de 5 porcento por an^.^^ Entzo, dos 54 desigualdades ngo aumentem de forma significativa, os indices de pobreza continuar5o a 29 As taxas de crescimento silo mais elevadas para 2003-2004 e tambkm para o periodo de previsilo de curto prazo 2005-2006. Depois disso o crescimento a longo prazo situar-se-a em tomo dos 5 porcento. xlvii MEMORANDO ECONOMICO PA~S DO porcentos registados em 200213, o indicede pobreza poderi reduzir at6 40 porcentos em2007. E possivel que o nivelde pobreza alcance o Objectivo de Desenvolvimento do Milenio que visa reduzi-la emmetade entre 1990 e 2015. (Se, por outro lado, as desigualdades aumentarem de forma acentuadacomo na maioria dos paises registando rapido crescimento, entgo o indice de pobrezabaixark menos lentamente.) Resumo das recomendaqaes clxvii. 0 conjunto das recomendaqties e apresentadono corpo principal do relatbrio. Iremos aqui, por questties de clareza, destacar as recomendaqbes-chave. A primeira e mais importante e agestio macroeconbmica: Estabilidade macroeconbmica, Sem inflaqiio baixa e control0 fiscal, nenhuma das reformas estruturais tera sucesso na manutengiio de boas taxas de crescimento e reduqgo da pobreza. Tomando em considerando os planos do Governo de aumentar os gastos relacionados com a pobreza, a 6nica forma de se contomar o defice e aumentando as receitas persistindo com a reforma fiscal, alargando a base tributaria e criando uma agencia fiscal independente. A gestcio dasJinanqas publicus deve ser melhoradaporqueMogambique encontra-se entre os paises com o mais fraco desempenho em Africa e sem melhorias substanciais, a dadiva dos doadores que tem apoiado o crescimento do pais sera comprometida. 0 SISTAFEdeve ser implementado na integra, os fundos extra-orqamentais incorporadosno orqamento, as cifras de execuqiio orgamental apresentadas de forma Clara e escalasalarial dos funcionarios p6blicos descomprimida. Os esforqos na area de eficigncia dos negocios: A fiabilidade de electricidade deve ser melhoradae o acesso a rede aumentado umritmo mais rapido, atraves da reestruturaq2ioda EDM. Na area dos recursos naturais: 0 acesso a hgua pelas populaqties rurais deve ser expandido realocando o financiamento dos sectores menos prioritarios. clxviii. Duas recomendaqbes adicionais emergiram da analise. IEessencialqueoGovernoprossigacomasuaestrategiadecombateao HIV/SIDA. Moqambique C umdos paises mais afectados em Africa, com uma taxa de prevalencia de 16,2 porcento. 0 custo a economia 6 na ordem de 1ponto percentualdo PIB per capita por ano. I0crescimentonaagriculturacomeqouaabrandardepoisdoatingiropic0no pos-conflito. Haver&necessidade de se introduzir mktodos de promog2io da produtividade. Estes incluiriioprovavelmente pesquisa melhorada que seja relevante ao agricultor, uma maior disseminaqBo de variedades resistentes a seca, o incentivo de mecanismos de subcontrataqgo de produggo, estimular a criag2io de associaq6es de agricultores, exploraq5o de novos modelos de acesso aos mercados financeiros pelos pequenos agricultores e facilitaq8o de investimento pelos agricultores de media e grande escala. xlviii CHAPTER 1. INTRODUCTION The context: a dramatic post-conflict recovery 1. Mozambique has staged a dramatic recovery from the damage o f the civil war. Since 1992, infrastructure has beenimproved and i s now approaching its pre-war levels, and incomes have risen considerably. The poverty headcount fell from 69 percent in 1996/7to 54 percent in200213. On average the economy grew by 8 percent annually between 1996 and 2003. This accomplishment can be attributed to the Government's, phased but determined approach to stabilization and structural reforms, as well as to concessional assistance (half o f Government expenditures), a remarkable agricultural "catch-up", an expansion o f agricultural exports, and fast expansion intourism, construction, and certain manufacturing subsectors. Another factor was the authorities' success inattracting "megaproj ects" in aluminium smelting, natural gas, and titanium mining, and a resultingtripling of exports inthree years. 2. Another factor which was a precondition for all o f the above i s the fact that Mozambique was successful inbringingabout reconciliation, endingthe civil war, and in managing potential conflicts since that time. The country had its third general and presidential election at the end o f 2004. 3. Inthe next three pageswe embark on abriefdescriptionofthe population and economy o f Mozambique so as to put the evaluative discussions o f the subsequent chapters incontext. A "snapshot" i s given o f the country's population, o f its still basically rural character, o f its poverty and social indicators inrelation to those o f its neighbors. This serves as the backdrop for the discussiono f the plan o f the Memorandumon page 4. A fast-growing population 4. Mozambique's population i s 19 million. Its population density i s one o f the lowest world-wide with 22 inhabitantsper square kilometer in2000. Evenby 2050, the population density will be still be low (Table 1). 5. The growth rate o fpopulation i s expected to fall from 2.4 percent (for the period 2000-2004) to 2.3 percent in2010. 30 The total fertility rate i s expected to fall from 5.7 in 2000 to 5.1 in2010, and to continue falling thereafter'to 4.5 in2020. Dependency rates are expected to fall from 90 percent in2000 to 83 percent in2010, as the working-age population grows more quicklythan dependants. Poor households are larger than nonpoor households, but the differential i s small (5.8 versus 4.5). 30 Klasen and Woltermann (2004), using UnitedNations Population Division (2004); andWorld Bank population estimates report slightly lower growth numbers, as they claim to take HIV/AIDS fully into account. MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Table 1. Demographic projections, 1990-2050 Population Pop. Population TFR Dependency Life expec- Lifeexpectan- (`000) growtha densityd ratee tancy (M)` C Y (*) 1990` 14 151 2.0 17.7 6.2b 92.5 42.1 44.8 2000 17 240 2.4 21.6 5.7 90.1 42.5 46.1 2005 19 420 2.4 24.3 5.4 86.2 45.2 49.0 2010 21 854 2.3 27.3 5.1 82.7 46.7 50.5 2015 24 518 2.3 30.7 4.8 81.2 48.7 52.7 2020 27 439 2.2 34.4 4.5 75.6 52.0 56.0 a,bThese refer to the annual growth rate in the 5-year intervalbetween the row where they are given and the following row. These estimates may not fully take into account the impact of HIV/AIDS. The life expectancy rates also refer to the annual life expectancy rates in the 5-year interval between the year of the row the numbers are entered and the subsequent one. Lifeexpectancy is in years at birth. The populationdensity is personsper square km. `The dependency rate i s the ratio of dependentpersons (below 15 and above 64 years) to the working age population,times 100. INE suppliesonly data for 1990-2000, without splitting into 1990-1995and 1995-2000. Source: INE web site, June 2005. 6. All these factors -barringthe likely impact o f HIV/AIDS -are mildly encouraging, suggesting that population growth inMozambique does not constitute a serious risk for poverty and inequality. While large household size generates an elevated poverty risk, that risk has declined over time, particularly inrural areas (Klasen and Woltermann, 2004). Various econometric exercises confirm the hypothesis that population dynamics are not a major driver o f changes ingrowth, per capita incomes, poverty, or inequality trends. Indeed demographic dynamics have helped support rising per capita incomes and falling poverty rather than hindering it. An overwhelmingly ruralcountry 7. Currently 73 percent3` o f the population i s rural, and with further migrationto the cities, this i s expected to fall to 71 by 2010. Rural-urban migration i s expected to continue. Projections indicate that between 2005 and 2010 the urban working-age population will grow by possibly as much as 4.1 percent per annum32- a rate substantially faster than that o f the rural working age population. This underscores the need for a growth path withjob creation, a point which will be raised inChapter 3 (p. 38) inthe discussionabout labor markets, job creation and poverty. 8. Approximately 80 percent o f the work force draw the bulk o f their income from agriculture, forestry or fisheries. Inaccordance with economic theory, the share o f the population involvedinagriculture i s falling. This structural change, along with others, i s discussed in Chapter 3 (p. 39). GDP arising from agriculture and forestry i s still the largest among the sectors, at about 20 percent. Given the weight o f agriculture inthe economy, it i s not surprising that agriculture, despite generally slower growth rates than the "modern" sectors, made the second-largest contribution to GDP growth inthe past decade, exceeded only by the manufacturing sector which included the so-called "mega- projects" (see p. 16). Therefore in Chapter 2 on growth, much o f the discussion centers 31 Source: INE (2000). As o f the 1997 census, 76.5 percent o f the population was rural. 32 See further discussion about population and the labor force, in the background paper by Benito-Spinetto and M o l l (2005). 2 CHAPTER1: INTRODUCTION on explaining the growth inagriculture hitherto and developing hypotheses about the likely drivers o f agricultural growth inthe future(see p. 17ff). A poor country, but lower poverty and inequality than the Africa average 9. Despite the country's recent progress, Mozambique i s still among sub-Saharan Africa's poorest countries. Gross national income per capita i s $250 ($1,060 inPPP terms) compared with the Africa average o f $480 ($1,750 inPPP terms). Consistent with its lower level o f income, Mozambique has a slightly lower share o f agriculture in GDP (19 percent versus 17percent for sub-Saharan Africa). Also consistent with its lower income, Mozambique i s less urbanized: 27 percent versus 36 percent for sub-Saharan Afixa. 10, This said, it i s striking that Mozambique has a smaller fraction o fpeople living below $1per day: 29 percent, versus 46 percent for sub-Saharan Africa. This i s because consumption (and by implicationincome) i s distributed more evenly in Mozambique, which has a Gini of 0.40, compared to the sub-Saharan Africa Gini of 0.49. 11. Interms of socialindicators-see Table 2 -Mozambique i s close to the average o f sub-Saharan African countries, but lags behindinliteracy, water access and HIV/AIDS: P infant mortality is 101per 1,000, close to the Africa average of 103; P life expectancy is 45 years, close to Africa's 46; P HIV/AIDS prevalence is higher, at 16.2percent, than the African meanof 7 percent. The implications for economic growth, and measures to combat the disease, are discussed inChapter 6 (p. 105). P gross primaryschool enrolment is over 100percent, betterthan the Africa average of 87 percent. Literacy, at 45 percent comparedwith Africa's 65, reflects the serious educational backlog o f the colonial and socialist periods; P water access, at 37 percent, falls far short of Africa's 58 percent. This ismainly because rural water access i s low, at 27 percent (Table 10, p. 26), behindAfrica's 46.5 percent. Expanding ruralwater access i s discussed inChapter 5 on natural resources (p. 75). 3 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Table 2. Economic and social indicators for Mozambiqueand sub-SaharanAfrica Indicator Mozambique Sub-SaharanAfrica Population (millions) 19.4 703 Populationgrowth (annual, percent) 2.4 2.3 GrossNational Income per capita (Atlas method, US$) 250 O 480 Gross National Income (PPP terms, US$) 1,060 1,750k Agriculture as percentageof GDP 19.4' 17 Poverty (?Aofpopulation below USSl/day in PPP terms) 29 46.4 a Gini index of inequality (96) 40.0 48.6' Urbanpopulation (96 of total population) 27 36 Life expectancy at birth (years) 45 46 Infant mortality rate (per 1,000 live births) 101 103 Prevalence o f HIVIAIDS (%) 16.2 7.2 Access to an improved water source (% ofpopulation) 37 e 58 of which: rural (% ofpopulation) 27 e 46.5 Literacy (% ofpopulation aged 15+) 45 a 65 Grossprimary enrolment (% of school-age population) 110' 87 Male 121 94 Female 100 '' 80 Source: unless otherwise indicated, from World Bank, Mozambique at a Glance, in tum from the Development Data Platform, which is compiled from avariety of sources. *From World Development lndicators 2005, p. 67f. Datarefer to the year 2001. This is the percentage of people living on less than 51.08 aday at 1993 internationalprices (equivalent of $1 .OO at 1985prices). See source at footnote 31 (p. xxxiii). Population growth in 2005: see source in Table 1, p. 2. Lifeexpectancy:see source in Table I,p. 2. Water: source is Fox et ai. (2005), Table 5 (on access to services). Primary gross enrolment: from World Bank (2005a, Table 3.3,p. 21), in tum from Ministry of lived on less than US$I(PPP terms) per day in 2002/3. HIV: WorldDevelopment Indicators 2005, p. 118, Table 2.18. ' Education,EducStnt annual schoolsurvey, 2003. Fox et ai. (2005)report (p. 4) usingIAF data that 29 percent of the population Agriculture share of GDP: WorldDevelopment Indicators 2005, p. 204. Datarefer to 2003. 'Agriculture share of GDP for Mozambique: see Table 6, p. 15. Data refer to 2003. GNl, PPP: WorldDevelopment Indicators 2005, p. 23f. Data refer to 2003. ' 'Gini: World Development Indicators 2005, p. 74. Simple average. Medianis 47.9. Datarefer to various datesbetween 1997and 2002. Gini for Mozambique: from IAF 200213, see Table 18,p, 35. " From IAF 200213, from underlying computations for Fox et al. (2005); refers to populationaged 18-65.O 2004. Rural water: World Bank WorldDevelopment Indicators Database (htiu..,de\d~iu.~iorldbank.or~'cxtcnial~drrcomn~aso),from: World HealthOrganizationandUnitedNationsChildren'sFund, inturn Global Water Supply and Sanitation Assessment 2000 Report. 12. There are enormous challenges ahead. Mozambique's infrastructure i s inadequate -especiallyruralwatersupply-andthereareseriousunmeteducationandhealthneeds. Manyo f the "first-generation" reforms associatedwithmarket liberalizationhave already been implemented. The country now faces the prospect o f tightening macroeconomic constraints, a need for substantial institutional improvement to make growth sustainable, and an increasingneed for better prioritization and management o f public expenditures to eliminate absolute poverty. Objectives and structureof the Country Economic Memorandum 13. Objective. This Country Economic Memorandum (CEM) has examined the growth-poverty linkage, usinga wide variety o f data sources, including the recently completed national household survey (2002/3) and the Demographic and Health Survey (2003). It has sought to understandthe sources o f growth inthe recent past, to evaluate the prospects for growth inthe next decade, to examine the likely implications for poverty, and to outline the policies that will be needed to achieve further growth and poverty reduction. The findings and recommendations are presented inthe following chapters: 3 The growthrecord 4 CHAPTER1: INTRODUCTIO P The povertyrecord (including linkages between growthandpoverty) P Challenges to growthandpoverty reduction: private sector development P Challenges to growthandpoverty reduction: naturalresources P Growthandpoverty outlook 14. Structure. This Country Economic Memorandum does not attempt to provide even coverage of all issues of economic policy. The previous CEM (World Bank, 2001d) provided a bird's eye view covering all major aspects o f the economy. This CEM i s focused on the growth-poverty linkage. There were two reasons for wishingto focus on this at this time. The first is that the Government was planningto develop its second Poverty Reduction Strategy Paper (termed PARPA inMozambique) and it was felt that a substantial contribution inthe area of growth and poverty might make a useful contribution to this. Second, the Government hasjust completed the preparation o f the second national household survey (2002/3). Thus for the first time data were available to examine the trend of poverty. Hence it was considered important to seek to explain in detail what hadhappened to consumption, employment, and poverty, and explore the linkages with the growth process. 15, Hence the backbone of the CEM i s first an examination o f the factors that will help to sustain growth (Chapter 2) and engender further poverty reduction (Chapter 3) inthe future. This approach implies certain specific limitations. The key sectors o f importance for growth and poverty reduction inthe future, particularly agriculture, miningand tourism, take pride o f place. Another limitationi s that the important subject o f corruption inthe public sector is not covered. In2004 the Government carriedout a surveyof perceptions o f corruption, with assistance from the World Bank, and undertook in2005 to publishthe results. A further limitationi s that Chapter 3 on poverty has as its main subject the impact of growth onpoverty. It i s not intended to cover the vast subject The impact ofpublic policy onpoverty; since this would requireapublic expenditure review, it was decided to postpone this for future work. 16. Inaddition-on account ofthe light coverage the subject hasreceivedinother sector work, Bank-produced and otherwise -the CEM also examines the relevance o f natural resource management to growth and poverty objectives (Chapter 5). It was found, inthe event, that natural resources are o f great importance for both growth and poverty reduction. Hence it i s hoped that inthe future the natural resource issues will be fully integratedwith the national policy discussions. 17. After the preparationo fthe CEM had begun, the Government made an additional request, namely that it include a chapter on private sector development. This had not formed part o f the original plan, because a great deal o f valuable research had already gone into this important subject (as i s documented in Chapter 4, p. 47). Hence it was agreed that the contribution o f this Memorandum would not be to break new ground in the area o f private sector development, but to arrive at a workable synthesis and prioritization of the many disparate recommendations arising from these efforts. Hence Chapter 4 is not the all-encompassing review o f all private sector- and business-related policies found inmany Country Economic Memoranda; insteadit focuses on a small number ofkey administrativebarriersto growth, and also presents a matrix with suggestedprioritized actions. 5 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM 18. Finally, while the CEMwas beingproduced, four other papers inprivate sector development and agriculturalpolicy were under way. To avoidduplication, it was decided to report only their mainpolicy conclusions inthe maintext, andto add brief summaries inthe form of boxes. These studies are: Growth and competitiveness (Box 1, p. 51), Rural development strategy (Box 2, p. 112), Impact of extension services (Box 3, p. 113) and Contractfarming and supply chainjnancing (Box 4, p. 114). 6 CHAPTER 2. THE GROWTHRECORD A. INTRODUCTION 19. This chapter surveys Mozambique's progress increating and sustaining growth over the past decade. The key threads o f the inquiryare to ask three questions: 3 Was growthofan enclave natureor was itbroad-based? 3 Are the sources o fgrowthobserved inthe past decade durable, that is, canthe same mechanisms be relied on in the future? Was part o f the growth w e have seen due to a one-off, post-conflict boom? 3 Didgovernment efforts assist inthe growthprocess? 20. The chapter starts with a briefreview o f the main data sources and the macroeconomic background including trade policy, fiscal policy and public financial management. It reviews the prospects for population growth. Then it examines the growth process from four different perspectives: > the demand side > the supply side (sectoral contributions) 3 a Solow-Denison growth accountingexercise, and > idiosyncratic contributors to growth covering, inparticular, the influence o f the mega-projects and o f foreign aid. Different methodologies are used inthis examination and this may well constitute a strength o f the analysis, given that there are certain common elements that emerge that serve to corroborate the key findings. A concluding section summarizes the main findings and attempts to answer the questions listed above. B. GROWTH AND MACROECONOMIC FACTORS 21. Mozambique's growth averaged a spectacular 8percent from 1996 to 2003 (Table 3), drivenby megaproject construction, investment from neighboring countries, buoyant donor support, and healthy agricultural The 2000 floods reduced growth to 1.5 percent but the economy recovered quickly, growing at 13 percent the following year. 33 See also more detailedmacroeconomic variables inAnnex Table 1, p. 135. MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM Table 3. Basic macroeconomic indicators 1997 1998 1999 2000 2001 2002 2003 2004 P e r c e n t a g e , u n l e s s o f h e r w i s e s t a l e d !Realannual GDP growth rate 11.1 12.6 7.5 1.9 13.1 8.2 7.8 7.2 /NominalGDP (Mt trillions) 39.8 46.9 51.9 58.4 76.5 96.9 114 137 /NominalGDP (US$billions) 3.45 3.96 4.09 3.72 3.70 4.09 4.79 6.09 IInflation (period average) 6.4 0.6 2.9 12.7 9.0 16.8 13.4 12.6 Gross domestic savings / GDP 8.1 10.8 13.7 11.6 8.0 11.0 10.8 11.8 Investmenti GDP 20.6 24.2 36.7 33.5 25.9 29.8 25.9 20.1 Governmentrevenue/ GDP 11.6 11.3 12.0 12.9 12.4 12.4 12.9 12.3 Govt. expenditure / GDP 23.9 21.6 24.7 26.6 32.1 30.0 26.5 23.7 Aid i GDP a 15.6 13.4 13.3 21.0 16.7 16.1 15.7 13.2 Interestrate (commercial lending rate) 24.4 19.6 19.0 22.7 26.7 24.7 22.0 Current accountdeficit/GDP (excl.grants) -17.7 -18.9 -28.2 -27.2 -26.1 -23.1 -19.9 -13.8 Exchangerate (000 Mt /US$) 11.5 11.9 12.7 15.7 20.7 23.7 23.8 22.6 NPV externaldebt/exports (percent) 709 549 212 177 110 92 102 84 Sources: Live Data Base, in tum from Mozambicanauthorities, IMF and World Bank staff estimates. a Aid includesgrants and public sector borrowing. 22. There i s no room for complacency. Over the mediumterm, high growth rates, in particular inthe agricultural sector, could beput at riskby constraints to further improvements inproductivity, such as land rights, transportation costs and volatile internationalprices, and in addition there i s the certainty o fperiodic droughts and floods. 23, Despite theprogress made, there is still a danger of macroeconomic instability. Reorganization of the banking system inthe mid-l990s, together with tight money, resultedin single-digit inflationup to 1999. Butthere have beentwo distinct phaseso f inflationbrought on by lax monetary policy, one starting in2000 and resulting in inflationo f 17 percent in2002, and another in200314 resulting ininflation o f 15 percent inearly 2004. Inbothinstances, monetarypolicy hadto betightenedby increasing reserve requirements and/or raising the bank rate. One o f the costs of this macroeconomic uncertaintyhas beenextremely highinterest rates, e.g. primerose to 35 percent at end-2002, though it has fallen since.34 There are highcosts for the small and medium-sized firm sector; fortunately duringthe periodunder review much o f the growth was drivenby large firms not dependent on Mozambican financial markets. 24. Mozambique's exchange system i s a managed float. Commercial banks may buy and sell foreign exchange to individualcustomers on a fully negotiable basis. The Banco de Moqambique sets a daily rate, which i s adjusted gradually inresponse to significant restrictions on the making o f payments and transfers for current account transaction^^^ deviations relative to the average selling rate o f banks at exchange bureaux. All have beeneliminated, with the exception o f the needfor prior approval for remittances of family living expenses over $5,000 per month, which i s granted by the central bank on a 34 An allied problem i s a persistentand large spreadbrought on by a lack of effective competition inthe banking sector combinedwith large bad loanportfolios intwo of the biggestbanks. 35 InIMF parlance, Mozambique i s an Article XIV country (IMF, 2003a, Appendix I, 2.) Prior approval p. mustbe obtainedfrom the centralbank for payments of invisibles above US$5,000 as well as for transfers o f profit remittances abroad. The possibility o f accepting the obligations under Article VI11i s being studied. 8 CHAPTER2: GROWTH case by case basis. Ineffect the latter restrictions i s not bindingbecause individuals may also purchase foreign exchange privately. Exporters may retain foreign exchange earnings infull and dispose o f them at will.36An auction system for foreign exchange will beput inplace in2005. 25. The real effective exchange rate had a major appreciation-and-depreciationinthe mid-1980s. Since 1992 it has fluctuated within a band o f k16 percent (Figure l), the and 2004 figure i s almost exactly the same as that o f 1992 (index = 54), which inturnis considerably depreciated compared with 1980 (index = 100). Eventhough aidpayments have amounted to 12-19 percent o f GDP since the 1990s, only inlate 2004 did there emerge any appreciation o f the currency, and this was short-lived as the nominalrate returnedto its previous level in2005. Onthe face of it there does not seem to be a"dutch disease" problem. 350 - -_____I_ _____I 150 _ 0 1980 1983 1986 1989 1992 1995 1998 2001 2004 Figure 1. The real effective exchange rate (1980=100, d e ~ r . ) ~ ' - Trade policy 26. Mozambique has made important progress in tradepolicy. The trade weighted average tariff i s 9 percent (Yagci, 2004), one o f the lowest inAfrica. There are four non- zero bands. The top tariff rate has beensteadily lowered, from rates approaching 100 percent inthe early 1990s, down to 25 percent in 2003. Furtherreductions intariffs will follow as the SADC trade protocol i s implemented. The plan i s to reduce the top tariff rate to 20 percent in2006. Management o f customs was contracted out duringthe 1990s, leading to increases inefficiency o f collection which more than compensated for the decline inthe tariff ratesS3* 27. Mozambique's export performance has been strong, growing at 22 percent annually (inUSdollar terms) between1996 and 2003, mainly owing to mega-projects. This expansion is substantially faster than that o f world exports o f about 6 percent. As a result, Mozambique i s one o f the rare countries inAfrica whose share inworld exports has risen. However, the "traditional" export base i s still weak and poorly diversified (mainly agricultural), having grown at only 2.3 percent inthe same period. (See Table 25 (p. 57) for detailed composition of trade data.) 36 The export surrender requirement was abolished in 1997 (IMF, 2003b, Annex, p. 4). 37 Source: IMF. A decrease (e.g. 100 in 1980 to 49 in 1995) denotes a depreciation. 38 See hrther discussion of trade in Chapter 4 on private sector development, p. 46. 9 MOZAMBIQUE COUNTRYECONOMIC MEMORANDUM 28. Inapoor economy like Mozambique, the expansion o f exports has to constitute a key avenue for growth, because local purchasing power i s not sufficient to support strong growth rates. Inaddition, emphasis needs to be placed on attracting direct foreign investment because the national saving rate i s low. 29. Market access i s not a bindingconstraint on export growth at the present, as the country has access to the EUmarket under the Cotonou Agreement and the Everything- but-Arms Initiative, and to the U S market under the Africa Growth and Opportunity Act. The country's access to the SouthAfrican market is also relatively free. Inthe medium term, market access may be a bindingconstraint when the country reaches its full export potential. To avoid this, Mozambique needs to participate inthe on-going mid-term review o f the SADC Trade Protocol to improve the rules o f origin, and also inthe Economic Partnership Agreement with the EUwhich will replace the Cotonou Agreement. 30. "Behind-the-border" problems continue to be the chiefbarrier to the needed expansion o f "traditional" exports. Business startup i s still expensive and slow, and contracts are difficult to enforce through the legal system. Stimulating growth inexports will also require improvements incompetitiveness by addressing labor rigidities, high transportationcosts, and highlending risk. Further priorities are: the extension o f export incentives to small companies, accelerating VAT reimbursement, setting up an export credit system, and improving standards and capacity to meet the sanitary and phyto- sanitary requirements, particularly for the EUmarket. Further detail o n trade policy is given in Chapter 4 on private sector development, page 47. Resource mobilization 31. New tax laws have improved revenueperformance, but much remains to be done.39 With a view to long-term fiscal sustainability, a value added tax was introduced in 1999 and is now the largest single taxationitem. A large taxpayer unitwas launched in2001. A new income tax law was passed in2002, rationalizing corporate andpersonal income taxes, reducing the corporate tax from 35 to 32 percent, and broadening the tax base. A new code o f fiscal incentives was passed, establishing standard concessions and transparent rules for foreign investors. The plan o f the PARPAwas to raise revenue from 11percent o f GDP in 1998 to 15.4 percent o f GDP by 2005; inthe event, revenuerose to only 12 percent in2004 (Table 3). It would have risen further had it not been for a slackening o f the revenue effort in2004, possibly associatedwith the elections inthat year. A set of administrative measures were adopted in early 2005 which the Government i s confident will enable the PARPA goals to be met with a delay: 15 percent by 2008. 32. Dejcits after grants were relatively low until 2000, thanks to a prudent fiscal stance accompanied by substantial external assistance. Combined with a prudent monetary policy, especially inthe period 1996 to 1998, and a program o f structural reforms based on privatization, tax and customs reform and trade liberalization, this resulted in low inflation, highprivate investment and highgrowth rates. Since 1998 there has been a shift inresources infavor o f health, education and agriculture, reflecting an 39 For a thoroughreview of tax inMozambique, see Coelho et al. (2001). 10 CHAPTER2: GROWTH increasing anti-poverty focus. Education, health and agriculture increased their combined share intotal budgetary allocations from 29 percent in 1998 to 39 percent in2001, where their share has remained. 33. However, Mozambique remains aid-dependent, with deficits before grants rising from 11percent o f GDP in 1998 to 19percent in2002 (Table 3). Expenditures grew at 17 percent annually in real terms from 1997 to 2002, due to bank bailouts, an increase in the civil service wage bill o f 46 percent inreal terms between 1999 and 200240, and to higher social spending. These deficit levels have only been possible because of high levels o f foreign grants. External assistance i s likely to remainhighinthe short run, but inthe longrunMozambique is unlikelyto be able to hrther increase its share intotal world aid; given that the latter i s growing by about 2 percent annually, and the Mozambican economy i s likely to grow at more like 5-7 percent annually, this would imply a gradual decrease o f the ratio o f aid to GDP from 15 percent currently to 11 percent by 2015. Hence it i s appropriate that fiscal adjustment has become a priority o f Government policy. It i s also appropriate that further efforts be made to improve the prioritization and efficiency o f expenditures. 34. The macroframework envisages important changes on the revenue and spending sides. Table 4 presents a feasible scenario, on the basis o f reasonable assumptions o f economic growth, revenue capability, and donor contributions (with grants averaging some U S 3 5 0 million annually). Note that here the intention i s to look at the growth numbers that the authorities anticipate. Since these (obviously) do not incorporate meteorological or institutional risk, they are on the optimistic side, but this may not be inappropriate for the short to medium term. (See the less rosy longer-term projections Chapter 6, p, 103.) Revenue i s programmed to reach 13 percent o f GDP in2005. Inthe pursuit o f long-term fiscal sustainability, revenue should reach 16-17 percent by 2010. On the expenditure side, spending is programmed to fall, inline with the medium term perspective o f the Government's Action Planfor the Reduction of Absolute Poverty (PARPA), from the very highlevel o f 32 percent o f GDP in2002 to 26 percent in2005 and 24 percent in2010. This represents a considerable reduction inpercentage terms but inreal terms expenditures are still growing at 4 percent per year between 2002 and 2010. Hence it i s not politically unrealistic. 40 Thus the billgrew by 13 percent per year between 1999 and 2002, well inexcess o f GDP growth. 11 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Table 4. Macro framework, 2003-2006 2003 2004 2005 2006 2007 Actual Est. P r o j e c I e d Output and prices Real GDP growth ratea 7.8 7.2 7.7 7.4 6.4 Inflation(period average) 13.4 12.6 8.0 7.3 6.5 Exchangerate (000 MtivSS) 23,782 22,581 ,. Money f a s a p e r c e n t a g e o f G D P ) Money and quasi-money 28.3 24.9 24.5 24.3 24.3 Net foreign assets 12.7 17.5 15.2 13.8 13.5 Net domestic assets 9.5 6.7 9.1 11.2 12.4 Credit to the government -3.1 -3.1 -0.6 0.8 2.1 Credit to the rest o fthe economy 12.6 9.8 9.7 10.3 10.2 PublicFinance Total revenue 12.9 12.3 13.2 14.0 14.6 Total expenditure andnet lending 26.5 23.7 25.6 24.4 24.5 Current 14.4 13.8 13.7 13.7 14.0 Capital and net lending 12.1 9.9 11.9 10.6 10.5 Overall balance before grants -14.0 -11.7 -12.4 -10.2 -10.0 Primary balanceafter grants -3.3 -3.4 -5.2 -3.5 -3.3 Publicexternal debt NPV pub.externaldebt/exports(percent) 102 83.8 83.6 89.5 96 Sources: Mozambicanauthorities;World Bank and IMF staff estimates andprojections. 35. The public external debt, as was remarked above, was considerably reduced through the two HIPC operations. Combined with the country's excellent prospects for exports, the net present value o f foreign debt service divided by exports i s projectedto fall from its current (2004) level o f 102percent to 74 percent by 2006 and to 50 percent by 2020. Mozambique i s well clear o f a growth-reducing debt overhang. Public finance management 36. Mozambique's fiduciary accountability can at best be described as weak, although the Government i s taking measures to improve it. The weakest areas are accounting, auditing and procurement. Mozambique scored 4 out o f 16 by one assessment4' and 5 out o f 30 by another - less than the sub-Saharan African average.42 This is a matter of concern inits own right. It will also be critical for growth inthe next decade, because: 41 The HIPC AAP rating, viz. Highly IndebtedPoor Countries initiative: Activity and Action Planfor expenditure reform (World Bank, 2003d). See footnote 42, p. 12, for the sort o f questionsasked. 42 The combinedWorld BanklEUiDfID assessment framework (Scanteam, 2004). 5 in 30 refers to the fact that there were five questionswhich scored"B", and none which scored"A", on a scale o f "A" to "D", "A" being good and "D" being puor. The questions cover: fiscal deficit, composition of expenditures,revenue outturn, arrears, risk oversight, budget coverage, budget classification, identification of poverty expenditures,audits, medium term expenditureframework, parliamentary review, cash flow management, debt management, intemal controls, payroll controls, procurement,timeliness of audits, and so on. 12 CHAPTER GROWTH 2: P improved spending efficiency i s neededto achieve better service delivery, which inturnis neededto spur growth, particularly ininfrastructure, education and health. It has been found that public works are substantiallymore expensive (20 to 30 percent higher) inMozambique than inthe region,43and so it i s incumbent upon the authorities to ensure that the most efficient procurement practices are being followed; P total expenditure can no longer expand at the rate it usedto. Total expenditures are likely to remain at around 26 percent o f GDP. Hence it will be essential to secure efficiency improvements inthe future; P donor fundinghas assisted growth andpoverty reduction, andwithout steady improvement inpublic finance management some o f the donor fundingmay be put at risk. This is especially important becauseMozambique is receiving more aid than many other countries. The justification for such generosity i s that the aid i s being well used - as evidenced inhigh growth rates and big improvements in social indicators -but if any o f these i s let slip, Mozambique's special justification will fall away and its rather weak fiduciary record may well result in its aid levels falling. 37. The Government has embarked on a program of reform inthis area. It has followed the right sequencing, starting with a new financial management law in2001 and accompanying regulations in2002, which set the basis both for modem accounting procedures and for procurement reform. The electronic financial control system SISTAFE i s being piloted intwo ministries in2005. Next will be the rollout to other sectoral ministries during2006, It i s expected that this will soon enable more accuracy in accounting and permit greater timeliness in auditing, thereby facilitating parliamentary oversight. Then should follow the improvement o f the classifiers (program, functional, regional) so as to provide detail sufficient to permit analysis o f poverty impacts. Budget comprehensiveness i s another goal which will be facilitated by the SISTAFE rollout, though further work will be neededto get proper accounting o f donor-funded spending. 38. Inparallel, the regulations and institutions governingprocurement are being revised so as to promote competition, transparency and value for money. 39. A further important task is to coordinate the SISTAFE implementationwith the reform o fpublic sector, and inparticular to enable coincidencebetween the staff establishment held by the Ministerio da Administraggo Estatal and the pay list o f the Ministry o f Finance. Inso doing it will be necessary to eliminate supemumerary staff the Government's public expenditure review o f education in2003 concluded that as many as 20 percent of the staff inprimary education were not properly accounted for.44 43 This has beenfoundby Bank staff working with Governmentprojectsinthe areas of rural borehole drilling, urbanwater contracts, roadrehabilitationand schoolbuilding. See World Bank (2003b). 44 See the discussionon "ghost workers" inWorld Bank (2003b). 13 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM c. EXAMININGGROWTHFROMTHEDEMANDSIDE 40. W e n o w proceedto examine the growth record in a systematic way, from four different perspectives: the demand side; the supply side (sectoral contributions); a Solow- Denison growth accounting exercise; and idiosyncratic contributors to growth covering, inparticular, the influence o fthe mega-projects and o f foreign aid. Inthis section, then, we examine the demand side. 41. Table 5 presents the standard breakdown o f the demand side o f GDP. Important factors underlying the growth o f GDP are the increase ininvestment (whose share in GDP rose from 27 percentto 32 percent) and the improvement on the extemal side, as the growth o f exports far outstripped the growth o f imports. The share o f exports in GDP rose from 14 percent to 20 percent. It will be seen below (p. 23) that foreign investment associated with the so-called mega-projects i s at the heart o fboth o f these phenomena. Table 5. Growthof GDP components, 1992-2003 I 1992-95 1996-99 2000-03 A n n u a l g r o w t h r a t e s ' GDP 3.1 9.6 7.8 Resource balance Exports 5.3 15.6 28.4 Imports 3.0 12.5 7.0 Consumption 1.4 6.5 4.0 Public 2.2 14.3 10.2 Private 1.5 6.0 3.3 Gross domestic investment 19.0 16.5 9.7 Public 12.0 12.0 16.2 Private 28.0 26.4 8.1 Sources: Live Data Base, in turn from Mozambican authorities, World Bank staff estimates and IMF. aArithmetic average of the 4 annual growth rates for each period. Least squares growth rates. Used on account o f large year-to-year fluctuations of investment in this period. (The arithmetic average growthrates were 4.6, 13.0 and 10.1uercent resuectivelv.) 42. Owing to the pronounced increases ininvestment and exports (most o f which in turn originated in the mega-projects), GDP as a whole grew faster than private consumption. Between 1996 and 1999, GDP grew 9.6 percent annually and private consumption grew at 4.7 percent; between 2000 and 2003 the equivalent figures were 7.8 percent and 4.0 percent.45 The wedge between output and consumption i s likely to persist over the next five years as exports from the mega-projects are expected to continue to grow at highrates. 43, The growth o f consumption, which was well inexcess o f population growth, provided the basis for the observed reduction inpoverty. M u c h depends, o f course, o n the distribution of the growth inconsumption. Inthe next section the supply side o f GDP 4s Consumptionas measured by the household surveys was higher than these numbers. It i s not unusual to find that consumption by the national accounts shows faster growth than that measured by household surveys - this has been observed inIndia and the United States. See the discussion by Deaton and Kozel (2004). (See also the reference inChapter 3 on poverty, footnote 67 (p. 34).) 14 CHAPTER2: GROWTH i s examined, and this will give some insight into where and how the consumption increase was generated, noting particularly the role o f agriculture. D. EXAMINING GROWTHFROMTHE SUPPLY SIDE 44. The sectoral breakdown o f GDP i s presented inTable 6. 1996 1997 1998 1999 2000 2001 2002 2003 P e r c e n t a g e Agriculture 30.4 30.0 27.3 25.7 21.2 20.1 20.6 19.4 Fishing 4.0 3.9 3.0 2.5 2.4 1.9 1.5 1.4 Mining 0.2 0.3 0.3 0.1 0.3 0.3 0.3 0.3 Manufacturing 8.6 9.5 10.9 11.6 12.0 13.6 12.0 11.7 Electricity and water 0.5 0.8 2.0 2.8 3.0 3.2 3.7 3.9 Construction 6.6 6.6 8.3 7.8 8.8 8.1 7.9 11.7 Commerce 23.8 23.0 22.3 22.1 21.5 22.7 24.7 22.5 Restaurants and hotels 0.8 1.2 1.1 1.2 1.3 1.4 1.3 1.3 Finance and insurance 3.7 3.2 2.7 2.0 4.0 3.7 3.8 3.2 Real estate 3.9 3.7 3.6 3.1 2.6 2.2 1.8 1.7 Public a w n . & defense 2.6 2.5 2.8 3.6 3.6 3.7 3.6 3.8 Transport and commun. 8.6 8.9 9.2 9.6 10.5 10.8 11.3 12.2 Education services 1.4 1.5 1.8 2.5 2.8 2.5 2.4 2.5 Health services 0.4 0.4 0.5 0.8 1.1 1.1 1.0 1.0 Other services 4.3 4.2 4.1 4.4 4.9 4.8 4.1 3.5 /Total 100 100 100 100 100 100 100 100 ources:Ministry of Planningand Finance, and mission computations. Shares computed using current Mt. 45, Between 1996 and 2003 there was growth inconstruction, tourism, certain manufacturing sub-sectors (particularly where the mega-projects are represented, as well as food and beverages), transport, services, and certain agricultural sub-sectors. There was less progress in certain manufacturing sub-sectors such as textiles. Detailed data on growthrates by sector are provided in Table 7 for 1996-2003. Also, Table 7 gives the contribution to overall growth, decomposed by sector, for the period 1995 to 2003. 15 MOZAMBIQUE COUNTRYECONOMIC MEMORANDUM Table 7. Gross domestic product by sector: real growth rates (percent) 1995 1996 1997 1998 1999 2000 2001 2002 2003 Mean Contri- 95-03a bution` herallGDP(m.p) 3.3 6.8 11.1 12.6 7.5 1.9 13.1 8.2 7.8 8.6 8.7 Agriculture 17.0 8.7 9.5 9.5 6.5 -13.1 10.6 12.1 9.0 5.2 1.69 Fishing 2.8 10.1 2.1 -11.0 -2.1 4.8 0.6 1.1 4.4 0.1 0.05 Mining 20.1 -19.9 21.1 20.6 -6.5 59.6 10.8 28.7 3.9 16.2 0.04 Manufacturing: incl.mega-projects 7.8 18.9 31.8 14.4 14.7 15.1 34.7 8.7 13.6 18.9 2.27 excl.mega-projects 7.8 18.9 31.8 14.4 14.7 1.9 0.4 0.7 4.3 9.9 0.81 Electr. & water 6.6 18.0 37.9 279.0 78.3 -8.3 9.9 10.4 3.9 43.1 0.60 Construction 19.0 24.0 18.1 26.2 3.4 13.0 6.7 10.8 7.0 12.8 0.86 Commerce -0.3 -1.0 8.3 12.1 2.5 3.2 17.6 4.6 4.6 7.0 1.15 Restaur. &hotels -5.3 18.0 35.5 -2.1 5.4 6.8 4.0 5.1 9.3 8.4 0.11 Finance & insur. 3.5 -3.2 29.9 -17.4 -26.9 80.8 21.3 15.8 10.2 10.4 0.19 Real estate 2.7 4.4 2.3 9.8 2.8 1.5 4.6 0.7 6.1 3.9 0.15 Public adm. & def. -52.2 -5.6 -0.9 2.8 18.1 6.1 22.3 , 7.4 11.2 8.6 0.18 Tpt & comm. 11.1 10.1 17.3 4.8 9.0 2.6 6.9 8.4 14.4 8.2 1.oo Educationservices 10.0 10.1 7.2 7.4 9.5 9.7 19.0 4.7 6.6 9.6 0.15 Health services 13.4 9.1 5.1 9.5 17.1 11.7 9.0 5.7 6.9 9.8 0.04 Other services 16.8 9.7 9.2 9.7 10.0 18.3 6.8 3.3 0.1 9.2 0.33 ;ource: Instituto Nacionalde Estatistica, Mozambique website. - The annual least sauares growth rate for the period 1995 to 2003. IfXis the variable and t i s time, the growth rate r i s [exp(b*)- ] where b* is the estimate of b from the regressionIn Xr= a + bf The latter is derived by logging the compoundgrowth . :quationX, =X,(l+r)` and setting b = In( l+r). This is the equivalent of fitting an exponential trendline and has the advantage of :xploiting information from all years and notjust the two ends. The very swift growth in the rubric "electricity and water" (42 percent) i s due to the provision of electricity for the MOZAL iluminiumplant during the construction period 1997to 1999. Contribution to overall annual growth in the period, inpercentage points. LetX,, be real output, where i indexes sectors, t ndexes time (referringhere to 1995) and t+n refers to 2003, Then the contribution to growth of sector i is 46. Growtho f the tourist industryi s reflected in Table 7 in growth o f the sub-sectors restaurants and hotels (10 percent annually between 1996 and 2003) and construction (13 percent between 1996 and 2003). As would be expected, given the expansion o f educational, health and other services o f government, the growth o f the relevant sectors was very high, around 10percent. The manufacturing sector had annual growth o f 19 percent, mucho f it due to the mega-projects MOZAL Iand II.46 With the latter excluded, average growth was 10percent for the period, slowing sharply to between 0.7 and 4.3 percent between2001to 2003. 47. The rightmost column inTable 7 presents the contribution to growthby each of the sectors, inpercentage points. These numbers take into account the share o f each sector. The most important single contributor to growthwas manufacturing, with an overall contribution to growth o f 2.3 percentage points per year, ifthe mega-projects are 46 See the section on mega-projects, p. 23. 16 CHAPTER2: GROWTH included, and 0.8 percentage points ifthey are excluded. This points up the significant role played by the mega-projects, discussed later inthis chapter (see page 23). The second most important single sector i s agriculture, with a contribution o f 1.7 percentage points per year. Although agriculture grew at a slower rate than did the other sectors, it i s the largest sector still (22 percent of GDP) and so its contribution to growth was large. W e saw inthe above section that consumption grew considerably duringthe period; the growth contribution of agriculture i s one o f the important elements insupporting consumption growth. As will be seen inChapter 3 on poverty, this important result ties inwith the finding that the sectorwith the largest contribution to poverty reductionwas agriculture. 48. The next most important contributors to growth were commerce (1.2), construction (0.9)and transport & communications (1.O). The contribution of construction is understandable because the mega-projects required a great deal o f construction, as can be seen inthe jump inthe growth rate o f the sector inthe late 1990s. The contribution o ftransport and communications reflects, among other things, the national road rehabilitation and maintenance effort and the growth o f the telecommunications industry,particularly through cell phones. The contribution o f infrastructure i s discussed below (p. 20). Growth in agriculture 49. Inthis sectionabriefdescription will be given o fthe agricultural sector and the factors underlying its growth inthe past decade. Inaddition, the future prospects for the agricultural sector are covered in Chapter 6 (p. 110), and the rural land access is dealt with inChapter 5 (p. 72) on natural resources. Three other pieces o f work have recently beenprepared: Rural development strategy, Impact of extension services, and Contract farming and supply chainfinancing. These are summarized inboxes 2 (p. 112), 3 (p. 113) and4 (p. 114) respectively, and so only their mainpolicy conclusions are mentioned inthe maintext. 50. While the industrial and service sectors o f the economy have the potential to absorb increasing shares o f formal sector employment inthe long run, Mozambique remains largely an agricultural country. Some 80 percent o f the economically active population i s employed inthe agriculture, forestry, and fishing sectors. The service sector (including government) employed 15 percent o f the work force, with industry absorbingjust 5 percent. Hence inthe medium term further reduction inpoverty will depend crucially on progress inthe smallholder agricultural sector. 5 1. Smallholders inMozambique generally have limited access to capital, have little schooling, and are exposed to meteorological risk and price risk. While 94 percent o f rural households operate some land (mean 1.4 ha4'), only about a third sell any crop output, The most important commercial crops are basic food crops (maize, beans, cassava, rice) followed by cash crops (groundnuts and cotton). Incontrast, fruit crops (cashew, mango and banana) are sold by many but insmall quantities and are not major earners. 47 Perumalpillai-Essexand Loening, 2005. 17 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM 52. The household surveys o f 1996/7 and 2002/3 reveal substantial improvements in the living standards of smallholders. This improvement arose from various sources: agricultural growth, improvements intechnology, increased market integration, and increased generation o f nonfarm income. As these candidates are discussed, in the subsequent pages, it will become clear that sustained productivity growth has not yet been achieved and that much of the growth in agriculture i s ascribable to factor increases. 53. Growth and "bounce-back". Overall, agricultural GDP growthbetween 1992 and 1997 was 6.8 percent per annum, slowing to 4.6 percent per annum between 1997 and 2003. It i s no surprise that agricultural GDP grew fastest inthe period 1992-1997/8/9. As late as 1996 internalrefugees were still returning from the cities to their ruralvillages. Smallholders contrived to raise their maize yields from a nadir o f 160 kg/ha in 1992 to over 1,000 in 1999, thereby catching up with the pre-Independence (1972) peak (see Figure7 below, p. 111). The annual growth o f foodgrains from 1994to 1999was 16 percent; that from 2001 (after the flood) to 2004 was 5.5 percent (Figure 2). Other big contributors to growth were sugar and tobacco. Some o f the increase in output arose fiom area extension, particularly in the period 1993 to 1998, when sown area grew at a very swift 6.5 percent per year. The process of retrieving, between 1992 and 1999, what had originally been achieved, might well be called "bounce-back": an increase infactor inputs (labor,4* land, unimproved seed), with very little technical advance, due to a post- conflict recovery. 1.6 0.40 1.4 0.35 c 5 (D Lo 5 1.2 0.30 -5 .- 0.25 6 Lo 1.0 $j 0.20f '10.8 r I 0.15 0.6 0.103 0.4 +Cassava (rt axis) 0.05 9213 93/4 94/5 95/6 96/7 97/8 98/9 99/00 00101 '011'02 '02/'03 '03/'04 Figure 2. Food crop output, 1996/7-2003/449 54. Evidence on informal cross-border exports (particularly maize to Malawi) confirms the impression of substantial food crop output growth given by the national data. 48 The labor force in agriculture grew by about 1percentper annumbetween 1996/7 and 200213. See the discussionand figures infootnote 83 (p. 39). 49 Source: 199314- 200314: MADER - DINA - Departamento de Aviso Previo. See also the synthesizing noteby Sousa, Loening and Perumalpillai-Essex (2004). 1992/3 fiom Finney (2003), intum from Aviso Previo. 18 CHAPTER2: GROWTH 55. Cashew and cotton production both made impressivepost-conflict recoveries but the gains appear to have tailed o f f since about 1998. As Mauritianinvestors rehabilitated the old factories, sugar production has shot up to its 1970 level, aided by a variable tariff of some 90 percent, as well as exemption from sales tax and investment incentives. 56. Improvements in technology. Food crop production is still rainfed, done by hoe, and mostly without fertilizer or improved seed. Some modest advances have been made, such as increaseduse o f drought-resistant varieties and animal traction. A survey o f farmers to examine the impact o f agricultural extension (World Bank, 2005e) revealed that 43 percent o f respondents had introducednew varieties inthe previous five years if they had had extension advice, and 21 percent had introducednew varieties ifthey did not have extension assistance. This likely explains some o f the yield increases inkey crops such as maize (see Figure 7, p. 111below). Use of natural or chemical fertilizer rose from 4 to 9 percent o f households between 1997 and 2003. Cash cropping became more widespread - cashew cultivation increased from 28 percent o f households in 1996/7 to 37 percent in2002/3, cotton cultivation increased from 5 to 7 percent, tobacco from 2 to 4 percent. Many o f the smallholders who increased fertilizer use are tobacco farmers inthe ZambCzia area, respondingto the impulse from Zimbabwean farmer immigrants. However, this list o f modest advances cannot explain value added growth o f 5 to 7 percent per year; this reinforces the argument that factor mobilization was the key element responsible. 57. There i s some modest evidence o f increased trade and marketparticipation. The north, center and south o f the country used to be near-autarkic due to hightransport costs and poor communication. With improvements inthe quality o f the roads, long-distance trade i s occurring and there has been a tendency for prices to converge across sub- regions. Meanwhile, the fraction o f households engaging in agricultural production for the market rose from about 14 percent in 1996/7 to 26 percent in2002/3. 58. Increasedgeneration of nonfarm income. There are some indications o f modest increases innonfarm income. Off-farm labor by agricultural households became commoner: in 1996/7 only 7 percent o f households had a member engaged inoff-farm labor, but by 2002/3 this had increased to 16 percent. Households with a member working for Government or NGOs rose from 2 percent in2002/3 to 6 percent by 2002/3. These improvements were generated mostly inthe three southern provinces (18 percent o f the population), where the natural resource base is limited and the climate i s not conducive to agricultural production. Growthin manufacturing 59. Between 1995 and 2003, manufacturing value added increased by 10.1percent per annum (18.1 percent per annum ifthe mega-project Mozal i s included, which was better than GDP growth which averaged 8.4 percent annually. The share o f manufacturing inGDP was 9 percent in 1996, and since manufacturing growth exceeded overall GDP growth, its share rose to 10percent by 2003 (13 percent with Mozal). 60. Manufacturing sub-sectors. By far the largest manufacturing sector i s food, beverages and tobacco, with 74 percent o f the total sector's output (Table 8). It i s followed by non-metallic minerals with 16 percent. All the rest have shares o f 2 percent or less. 19 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Table 8. Real industrialproduction index by subsector (1999=100) 1999 2000 2001 2002 2003 Ann. Sharein growth' 2003 Manufacturing 100 109 121 144 203 19 100 % Food, beverages and tobacco 100 109 115 148 212 21 74 Textiles, clothes and leather 100 41 53 57 81 -5.0 1.6 Lumber 100 105 152 98 74 -7.4 0.3 Paper (graphic arts & publications) 100 85 92 110 130 6.8 1.7 Chemistry and petroleumproducts 100 102 111 89 92 -2.0 3.3 Non-metallic minerals 100 152 213 191 334 35 16 Metallurgy excl. Mozal 100 32 12 17 20 -33 0.1 Metallic products & tpt equipment 100 131 143 235 158 12 2.1 Other manufacturing products 100 127 141 160 189 17 0.03 Manufacturing growth between 1999 and 2003 was driven basically by the largest subsector, namely food, beverages and tobacco (21 percent annually inrealterms). Lesser contributions came from non-metallic minerals (growing at 35 percent annually), paper (7 percent) and metallic products (12 percent). The remaining subsectors stagnated or declined. Growth in infrastructure 61. Mozambican business people report that two o f the few aspects inwhich the business environment has improved are telecommunications and roads. While it i s not possible to quantify the impact that this had on growth, i t i s worth documenting the improvements made. 62. The telecom liberalization has proven to be a remarkable success story o f institutional change. As o f 2000, there were 85,000 land lines and 51,000 cellphones. Telephone charges were highand the quality o f the cellphones was poor. The government created an independent regulatory body, revised the sector law and attracted a new mobile operator. By 2005 there were 800,000 mobile phone subscribers, many o f them outside o f Maputo, and the quality o f service has improved substantially. 63, While Mozambique still has one o f the least developed road networks inthe southern African region, it can boast o f remarkable achievements innetwork rehabilitation. Between 1992 and 2000, roads ingood or fair condition rose from 10 percent to 57 percent o f the classified network, and impassable roads decreased from 50 percent to only 8 percent.50 This noteworthy achievement i s superior to that o f some o f Mozambique's neighbors. For instance, Mozambique's unpaved roads ingood to fair condition rose from 20 percent to 5 1percent, while those in one large neighboring landlocked country decreased from 65 percent to 26 percent. Given the importance o f 50 Since then, road ingood and fair condition rose to 70 percent by 2003, and impassable roads fell to 5 percent (source: the Mid-Term Review o f the 10-year roads program inM a y 2005). However, these data are less reliablebecause they are not survey-based. 20 CHAPTER2: GROWTH roads for internal trade, for communications, for agricultural production, and transport o f personnel, the rehabilitation of the road network may well have contributed to the growth process. E. A "GROWTH ACCOUNTING" APPROACH 64. I t i s usefulto try to understand the extent to which factor increases and productivity increases have played a role ininducing Mozambique's growth inthe past decade. Recent analysis by the IMF(2005) i s quoted at some length. A standard Cobb- Douglas production function i s used, together with assumptions which are commonly usedinthe growth accounting literature: (a) that the capital stock series canbe constructedusingthe perpetual inventory approach; (b) that depreciation was constant at 6 percent and that the capital-output ratio was 1.5 in 1980; (c) that population can be used as a proxy for labor force inthe absenceo f employment information; (d) that, as in Tahari et al. (2000), the labor and capital shares were 0.6 and 0.4 respectively. 65. Before proceeding to examine the results, the limitations o f the methodology should be stated plainly. Interpreting the "Solow residual" as total factor productivity i s a heroic assumption, becausebesides gains ineconomic efficiency, the residual may also reflect political factors, institutional and policy changes, climatic shocks and measurement errors (Bosworth and Collins, 2003, and Tahari et al. 2004). Ifthe residual i s only the "measure o f our ignorance" (Solow, 1957) then the TFP numbers are uninterpretable. 66. The results inTable 9 suggest that growth has beenincreasingly drivenby improvements in"total factor productivity" (TFP). This i s inline with similar exercises for other sub-Saharan African c~untries.~' The increases inthe TFP contribution in 1993- 2004 coincides with the postwardperiod o f macroeconomic stabilization and implementation of structural reforms and higher inflows o f foreign capital and concessional assistance. Table 9. Sources of economic growth from a Solow-Denison model (percent) 1981-2004 1981 - 1984 1985 - 1992 1993-2004 Real GDP growth 3.94 -6.0 3.2 7.8 Factor accumulation l / 2.4 -0.7 2.1 3.6 Solow residual (interp. as TFP) 21 1.6 -5.3 1.1 4.2 Memorandum items: Investment-GDP ratio 22.4 18.7 21.7 25.5 Real investment growth 6.2 -1.2 2.5 9.1 1/ Accumulation of labor and capital, using factor shares of 0.6 and 0.4 respectively. 21Residualfrom the growth accounting exercise. 5' See Tahari and others (2004) for Sub-SaharanAfrica countries' comparisons and Treichel(2005) for Tanzania. 21 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM 67. Factor accumulation accounted for most of growth inthe period 1985 to 1992 (2.1 percent o f 3.2 percent). Factor accumulationpicked up after 1993 to 3.6 percent per year, but TFP rose even more to 4.2 percent. Ifthe TFP findings can be believed, then Mozambique's experience confirms what the empirical literature has indicated are the main determinants of TFP growth: P improvedinstitutions and human capital development (health, education); P a favorable macroeconomicpolicy environment (less externaldebt, lower government consumption, lower inflation); and P diversification o fthe economic base arising from structuralreforms including privatization, financial sector reform, and trade liberalization. 68. We are on slightly firmer ground when restricting the examination to the factor accumulation side -because this frees us from the assumptions about the nature o f the TFP residual. The approach can be used, then, to examine the impact o f educationon growth. Fromregression approaches on household data sets, the impact o f education on consumption and on wages inMozambique may be characterized simply: P positive but small impacts (about 3 percent per year o f school) inrural areas - which reflects essentially smallholder agriculture; P large impacts (between 10 and 15 percent) inrural non-farm pursuits (Walker et al. 2004, and Perumalpillai-Essex and Loening 2005); P large impacts (5 to 15 percent per year o f school) inthe urban economy, with returns rising sharply at the higher levels (see Chapter 3, Annex Table 32. ,p. 158). 69. Carrying out the standard Denison (1967) decomposition o f growth yields two key findings (see Benito-Spinetto and Moll, 2005): P The rate o f growtho f labor force quality viz. the impact o f education on the - quality of the work force -i s 0.6 percent per year, within the range o f that achieved inother countries. Inorder to continue labor force upgrading at this pace, it will be necessary to vastly expand secondary schooling access. P InMozambique, education accounts for about 7 percent o f the growth observed between 1997 and 2002 (or about halfa percentage point o f growthper year). Again, this i s within the rangeo f what i s calculated for other countries. F. IDIOSYNCRATIC SOURCES OF GROWTH 70. What factors could explain Mozambique's idiosyncratically highgrowthrates? Candidates are: exceptionally highinvestment rates owing to megaprojects and links with neighboring countries; postwar catch-up and the recovery o f traditional markets, especially inagriculture, and massive foreign aid receipts. Investmentrates, foreign investmentand skills 71. As a result o fmassive aid, megaproject interest, and location adjoining South Africa, investment rates inMozambique were high. Investment as a percentage o f GDP 22 CHAPTER2: GROWTH rose from 23 percent in 1995 to 28 percent in2003. By contrast, investment during the 1990s insub-Saharan Africa was 19 percent o f GDP. 72. Foreign investment and skills. Duringthe 1990s, Mozambique became an attractive location for foreign investment. South Africa, inparticular, became the Mozambique's biggest trade partner, accounting for 40 percent o f Mozambique's imports and 20 percent o f its exports. South Africa has become the largest investment partner o f Mozambique. South African investment represents 35 percent o f foreign direct investment (FDI) inflows to Mozambique. Inmost years, FDIfrom South Afi-ica greatly exceeds domestic private investment. It involves such things as the Maputo Comdor (road infrastructure), Motraco (electricity supplyto Mozal), sugar, beer, soft drinks, cereal milling, most tourism facilities, and some agro-industry projects such as cashew processing and coffee. Post-war catch-up and recovery of traditional markets 73. Before independence Mozambique was the tourist's playground. Duringthe 1990s the country started to regain this large market, as well as attract international conferences. An indication i s the 17 percent growth rate in the number o f rooms in luxuryhotels. 74. Recovery o f traditional agricultural markets i s another factor. Inthe early 1970s cash crops accounted for almost two-thirds o f Mozambique's export revenues. During the socialist period, production o f cash crops declined sharply. Some agricultural markets are starting to recover their pre-independence volumes, e.g. citrus, bananas, sugar, and tobacco. Mega-projects 75. Mozambique currently has three "mega-proj ects"- the Mozal aluminum smelter Mozal, the Cahora Bassa hydroelectric plant and Sasol gas project which extracts gas at Temane and exports to South Africa via pipeline. Several more are planned up to 2010. Two factors weigh heavily inthe decisions o f these instances to invest inMozambique: tax and importation advantages which were negotiated on a one-on-one basis; and the availability o f natural resources - gas, titanium, water, and coal. 76. As a group, between 1998 and 2002, the mega-projects increased from zero to 7 percent o f GDP. By implication, the megaproject contribution to GDP growth inthe periodwas about 1.652percentage points o f GDP growth annually. On average, the megaprojects added some 0.4 percent o f GDP per year to fiscal revenues, and 48 percent o f GDP per year to exports o f goods.53 Contributions to the overall BOPare modest given the generated debits, such as the imports o f raw materials, debt service payments, profit repatriation, and investment and management services. Nevertheless, the mega- projects contributed on average US$113 millionper year innet flows to overall BOP 52 Divide the growth of GDP from mega-projectsby total growth of GDP: suppose GDP i s 1 in 1996, then total GDP growth is 1.0@-1,and total mega-projectgrowth is 0.07 1.0@. Dividing, the mega-project * contribution i s 0.07 * 1.086/ (1.086-1)= 20%. And 20% of the 8% growth rate = 1.6percentagepoints. 53 See Annex Table 8, p. 139, for a detailed table of the various impacts of the mega-projects. 23 MOZAMBIQUE COUNTRYECOKOMICMEMORANDUM between 1998 and 2002. It i s estimated they will contribute U S $ 327 million per year o n average between 2003 and 2010. 77. The mega-project direct contribution to local employment also varied greatly depending o n the phase the project was at (construction or operating). Mozal employed 5,000 people duringconstruction, and about 1,000 during the subsequent production phase. More important i s the contracting of Mozambican supplier firms which creates employment indirectly. Total local purchases by the mega-projects, at an estimated $50 millionper year, are still a fraction o f total sales o f $1billionper year (water, electricity, catering, cleaning and security), but the figure i s rising, with technical assistance from a multilateral aid project. 78. The tax contribution has been very limited - 0.4 percent o f revenues -given all the attractive incentives such as free profit repatriation under the provisiono f the industrial free zone legislation and the Investment Project Authorization. However, the authorities project that future contributions will increase to 3 or 4 percent o f revenues by 2010 as the special provisions start to expire in2005. 79. The future contribution o f the mega-projects, includingMoma Heavy Sands, Gaza Heavy Sands and the Mpande hydroelectric plant, i s expected to peak at 12percent o f GDP in2007, falling to 10-11percent by 2010 as construction terminate^.^^ They will contribute about 0.5 percentage points to the growth rate annually. All inall, their main impact on poverty i s through their tax contribution, and so a hitful avenue for poverty- reducing policy inthe future i s to simultaneously improve the efficiency o f tax collections and to improve the quality o f spending inpoverty-relevant areas. The megaprojects helped "put Mozambique on the map" and will probably encourage non- megaproject investment also. All these effects have beenpositive to greater or lesser degrees, and no negative effects have been observed. Hence overall there i s at least a primafacie case that the authorities' unorthodox policy o f tailor-made tax breaks and importation advantages has been successful.55 Foreign aid 80. Between 1997 and 2003, Mozambique's foreign aid receipts were o f the order o f 12-19 percent o f GDP (see Table 3, p. 8), or $500-700 million, including both program and project assistance. Mozambique i s one o f the continent's largest recipients o f aid, garnering $58 per inhabitant, more than double the sub-Saharan African average o f $26 (Figure 3). j4 See Annex Table 9, p. 140, for further detail on the future o f the mega-projects. 55 A complete cost-benefit analysis o f the mega-projects is beyond the scope o f this Memorandum. Such an investigationwould need to compare - in a general equilibrium framework - the cost o f the tax breaks and the poverty-reducing impact o f the lost taxes, versus the employment effects and the broadening o f the long-term tax base that the mega-projects brought. L4 CHAPTER2: GROWTH I 'IIou cf) UAid per capita (current US$) I Moram Mali Malawi Niger Uganda Ethiopia All sub- bique Saharan Africa Figure3. Selected sub-Saharan countries:Net aid as apercentageof GDP, and annualper capita aid, average 1983-200256 81, Duringthe periodunderreview, the natureof aid to Mozambiquehas changed in several dimensions. Program aid has gained inimportance vis-a-vis project aid, rising from 23 percent o f all aid in 1995/6 to 31 percent in200112. Grants have grown relative to loans, increasing from 52 percent o f all aid in 1995/6to 69 percent in200213. In addition, there were the two rounds o f the Heavily IndebtedPoor Countries initiative, in 1999 and 2001, which massively reducedMozambique's external debt stock. Sectoral allocations remained constant, barring health's share which rose from 9-11 percent in 1995 to 14percent o f all project aid by 2002.57 82. Didaid "work"? By and large it seems that it did. A broadrange of indicators in the health and education fields suggest substantial improvements since the early 1990s (Table 10). The key health indicator i s the infant mortality rate, which fell from 149 in 1995 to 101 in2003, one of the fastest reductions observed inAfrica. Ineducation, the headline figures are the gross enrolment percentage for lower primary, which rose from 56 to 100percent, and the completion rate for lower primary, which rose from under 20 percent in 1990 to 36 percent in2002. Literacy rose from 40 percent in 1996/7 to 45 percent in2002/3. There have beenbig improvements inthe quality and the coverage o f the road system, as mentionedabove. There have beenmajor improvements inaccess to safe water inrural areas - from 12 percent o f households in 1996/7 to 27 percent in 2002/3 and smaller improvements inurban areas. At the same time, it mustbe recognized that aid i s often used inefficiently. An example o f this i s given by the widely differing unit costs ofbuildingeducationalfacilities or rehabilitating roads, depending on which donor i s involved. 56 Source: World Bank, 2004. Africa Development Indicators, Chapter 12, tables 9 and 15. 57 See Annex Table 10, p. 140, for detail on the distribution o f project aidby sector. 25 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Table 10. Indicators: health, education, water, 1990-2003 Start year Statistic year End Statistic Health Infant mortality rate (per 1,000) 1995 149 2003 101 Use of contraceptivesby women (%) 1995 6 2003 17 Birthsattendedby ahealthprofessional(%) 1993 26 2003 48 HIV1AIDS prevalence(%) 1995 n.a. 2004 16.2 Vaccination coverage (%) 1995 55 2003 82 Education Gross enrolment rate, EPl(96) 1992 56 2003 110` Gross enrolment rate, EP2 (%) 1992 13 2003 35 Gross enrolment rate, ESGl (lower secondary) (%) 1997 6 2003 15 Completion rate, EP1 (96) 1990 <20 2002 36 Adults with completedprimary (EP2) (%) 199617 8 200213 11 Average years of education of the rural population: Men 1996 2.1 2002 2.7 Women 1996 0.7 2002 0.9 Literacy (self-ascribed, "can readand write") 199617 40a 200213 45b Urban (%) 199617 71a 200213 69b Rural PA) 199617 32a 200213 33 Men(%) 199617 59a 200213 64b Women (`A) 199617 24a 200213 30b Water Access of households to safe water, rural (%) 199617 12 200213 27 Access to householdto safe water, urban (?A) 199617 . 56 2002 64 Sources: Health from the two DHSsand the Ministry of Health. Educationenrolmentfrom the Ministry of Education. Average years of educationfrom the IAF. See further detail in Benito-Spinettoand Moll (2005). Water access from the IAF; see paragraph 270 and footnote 128 (p.80) for further discussionof water access figures. a Literacy 199617:from Ministry of Planningand Financeet nl. (1998), Table 2.10. In turn from IAF. Referring to adults aged 18-65. Literacy 200213:from IAF, underlying calculations for Fox et nl. (2005). Referringto people aged 18-65. Note that urban literacy has gone down- this is probablybecauseof the change in the definition of "urban" which resultedin the urban populationincreasingby half, and encompassingpeople with less educationwho hadpreviously been classified as rural. `GER: from World Bank (20051, Table 3.3, p. 21), inturn from Ministryof Education,EducSfnt annual school survey, 2003 83. What was the impact of aid on growth? Some simple approaches would suggest a substantial impact. An ICOR (incremental capital output ratio) approach suggests that aid may have delivered 2.5 to 4.0 percentage points o f growth per annum. Manipulating the coefficients relating to the effectiveness o f aid inBumside and Dollar's (1997b) multi-country study suggests an impact o f 4.8 percentage points. Alternatively, a study o f aid inpost-conflict societies (Collier and Hoeffler 2002) concludes that aid delivers an increment to growth of two percentage points inthe period 3-8 years after the reestablishment o f peace. 84. Inorder to investigatethe links betweeninfrastructure, public investment, aid, growth and poverty, a simple general equilibrium model was built, inthe spirit o f work by Agenor et al. (2004) and the well-known 1-2-3 model (Devarajan et al. 1994).58The 1-2-3 component o f the model posits a small open economy which produces an 58 Thanks to: R. AgCnor for the idea of linking growth, infrastructure, and aid (AgCnor et al. 2004), and for advice on modeling; D.Go for advice on modeling and for providing the 1-2-3 EXCEL-based model; L. Koryukin for EXCEL ModelBuilder. 26 CHAPTER2: GROWTH exportable and a non-tradable, and in which there i s imperfect substitution indemand as between the nontradable and imports, and imperfect substitution in supply as between the nontradable and exports. The 1-2-3 i s static, applying just to a single year. Inorder to get a multi-year perspective, a set o f growth equations i s set up inwhich output i s a function o f worker education levels, public health capital stock, population, the private capital stock, the public infrastructure capital stock, and the quality o f infrastructure. In addition, several variables are estimated econometrically. The details o f the model are explained at greater length inBenito-Spinetto andM o l l (2005). 85, Aid influences the economy through government consumption, government investment, and tax. Intum government investment i s allocated (by assumption, exogenously) over health and infrastructure investment, thereby adding to the existing stock of capital inhealth and infrastructure; and these feed into the equations determining the growth rate. The imperfect substitution assumptions permit usto examine the impact of policy changes on the real exchange rate. 86. The key simulation i s that o f a 10percent reduction inaid inthe first year, 2004. (This i s not to say that aid is about to fall; one might equally experiment with an increase inaidof x percent per year.) This results inareductiono f average annual growth between 2005 and 2010 by 0.42 percentage points. This would give an upper bound on the impact o f aid o f 4 percentage points.59 87. What happens to the real exchange rate inresponse to the simulated reduction in aid? If were to consider only the demand side, the lowered demand for non-tradables one would result in a decline intheir price, viz. result ina weakening o f the real exchange rate, But this model takes into account the supply side also. The withdrawal o f aid reduces public investment, particularly inhealth and infrastructure, so that public health capital and infrastructure capital grow by smaller amounts. This results inlower growth of output. Inthis case, the lessening o f supply neatly offsets the lessening o f demand for nontradables, and there i s barely any change inthe real exchange rate between the base and the simulation scenarios. Hence "dutch disease" appears not to be an important factor, This ties inwith the observation in Figure 1(p. 9) that the real exchange rate has strengthened by only 20 percent despite the massive inflows o f aid since 1992. Combiningthe sources of growth 88. Itis not possible, giventhe present state o f economic knowledge and given the country's data base, to give a complete breakdown o f Mozambique's growth rate by source. Nevertheless it i s possible, after the above investigation, to provide at least a rough explanation for a good part o f the growth observed. 89. There i s a broad overlap between the findings from the four dimensions o f growth -demandside, supplyside, growthaccountingandidiosyncraticfactors. Theincreasein investment and exports noted on the demand side (p. 14) corresponds to the large contribution o f the manufacturing sector on the supply side (Table 7, p. 16). This inturn corresponds to the "idiosyncratic factor" o f the mega-projects, which were calculated to have raised GDP by about 1.6 percentage points per annum (p.23). When looking at the 59 A CGE model (Tarp et al., 2002) indicated that, as of the mid-to-late 199Os, aid added 5 percentage points o f GDP per annum. 27 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM supply side, it was found that the agricultural sector accounted for the second-largest sectoral contribution to growth (1.7 percentage points out o f 8.6). This corresponds with the finding inChapter 3 on poverty that agriculture was the sector with the largest contribution to poverty reduction. The agriculture contribution also corresponds to the finding from the section on growth accounting(p. 21) that a substantial part ofthe growth observed was due to factor increases - inthe case o f agriculture, increases inland area. G. CONCLUSIONS 90. This inquiryhas shown a fairly optimistic picture about Mozambique's growth record. Inanswer to the first question - wasgrowth broad-based? - it i s clear that growth was broad-based, with considerable progress being registeredin agriculture, but also that there was growth invirtually all sectors, the sole exception being that parts o f the manufacturing sector declined (at least intotal employment - see Table 21, p. 39), probably due to the restructuring associated with privatization inthe mid-1990s. 91. Inanswer to the second question- is growth sustainable? -the response goes two ways. Some part o f the growth i s not sustainable. The bounce-back inagriculture i s unrepeatable and agricultural growth rates have fallen since their 1990s levels. The impact o f aid will decline inthe future simply because the amount of assistance will fall as a fraction o f GDP. 92. On the other hand, much o fthe growth observed is sustainable, provided there i s improved economic management. Mega-project investment appears to have good prospects for the next decade; tourism has good prospects also as will be seen in Chapter 6 (p. 116), and several key natural resources are far from fully exploited, as will be seen inChapter 5 (p.69). 93. Inanswer to the third question-did Government efforts assist in thegrowth process? - it has emerged that the authorities' efforts did indeedmake a contribution. Relative macroeconomic stability, the expansion o f infrastructure, and the development o f educationwere all essential elements. More effort will be needed, however, to secure improvements inpublic finance management and inresource mobilization. Inparticular, better focused attention i s needed for "behind the border" policies to facilitate trade, and this will be discussed further inChapter 4 (p.47). 28 CHAPTER 3. POVERTY A. INTRODUCTIONAND SUMMARY 94. This chapter surveys Mozambique's progress inpoverty reduction over the last six years. Using two cross-sectional national household surveys (1996/7 and 2002/3), growth inhousehold consumption, changes inthe distribution o f that growth and the role these two factors played inreducing poverty, are analyzed. Changes innon-income poverty measures are also analyzed, including changes in assets and access to services. The correlates o f poverty in2002/3 are analyzed usingbivariate and multivariate techniques. Finally, a profile o f household livelihood strategies and labor market behavior o f households i s provided as a basis for linking macro and sectoral strategies to households. 95, The main finding i s that as a result o f strong growth inincomes inthe agricultural sector as well as the non-agricultural sector, poverty declined rapidly inMozambique over the 96197-02/03 period in rural areas and inmost urban areas. The decline was broad based, and can be seen inimprovements in both monetary and non-monetary poverty measures. One key reason for the good poverty performance i s that inequality didnot change much, so aggregate growth inconsumptionreachedpoor households and raised their consumption levels. 96. Despite this good progress, more than 50 percent o f the population remains in poverty today. The poor inMozambique are mostly living inrural areas and working in agriculture, although increasingly one earner inthe household will get income from another sector as well. Poor adults have little education, and their children are less likely to be in school. Many still do not have access to safe water, and have inadequate housing. Lifting the other 50 percent out o f poverty will require continued broad-based growth in the economy, coupled with continued expansion o f social services to the poorest. B. TRENDS POVERTYMEASURES6' IN 97. There i s overwhelming evidence that income poverty declined sharply between 1996/7 and 2002/3, from 69 percent to 54 percent. Inthis examination, the Ministryo f Planning and Finance methodology was used, which calculates a consumption aggregate by deflating food prices temporally and spatially. " This amounts to computing different ' O This section builds on the analysis done in the Ministry o f FinanceandPlanning o f these two surveys, See Ministry of Financeand Planninget al., 2004. 6' An adjustment that was experimentedwith was to use adult equivalent consumption (that is, taking into account the size of the householdandcompensatingfor the fact that adults consume more than children) insteadof per capita consumptions. Inthe event none of the keyresults is changedby usingthis approach. MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM nominal poverty lines by province, by urbadrural location, and by year. Poverty fell in all provinces but Maputo and Cab0 Delgado (Table 11).62 Table 11. Poverty headcount by province (percent) 1996 2002 All 69.1 54.1 Urban 61.7 51.6 Rural 71 55.2 Niassa 69.9 49.5 Cab0 Delgado 56.8 62.8 Nampula 68.7 53.6 Zambezia 68 45 Tete 80.3 58.7 Manica 62.3 44.4 Sofala 88.2 34.1 Inhambane 83.8 81.1 Gaza 65.4 59.7 Maputo 64.8 71 Maputo city 47.3 53.2 Source Fox et nl (2005), usingthe IAF suneys 98. Rural poverty fell slightly more (by 22 percent) than urban poverty (by 16 percent), as emerges in Table 12. This result i s somewhat misleading because the Government changed the definition o f an urban area based on the census data, increasing the urban populationby 50 percent. Inthe last two columns, we show the changes using a consistent definition o f urban areas, and see that inthe most urbanized locations, poverty did fall instep with rural areas. 62 But see footnotes 63 and 64 (p. 3 1). 30 CHAPTER POVERTY 3: Table 12. Percentagechange in povertyrates, adjustingfor urbanhral border changes Using per adult equivalent consumption, consistent 1996 Using per capita consumption urbadrural Urban Rural Urban Rural Niassa -27.1 -26.3 -23.2 -30.3 Cab0 Delgado -15.5 15.0 0.6 11.3 Nampula -45.6 -11.6 -66.3 -9.7 Zambezia -23.1 -35.1 -58.1 -33.7 Tete -12.6 -29.3 -8.4 -28.9 Manica -14.9 -36.0 -7.3 -34.9 Sofala -44.5 -63.4 -47.1 -64.8 Inhambane 15.2 -3.0 18.1 -7.3 Gaza -17.4 -3.6 -14.1 -8.1 Maputo 28.2 5.5 33.1 7.4 Maputo City 12.8 12.5 All -16.5 -22.3 -19.4 -22.3 Source: Fox et al. (2009, using the IAF surveys. 99. Poverty reduction was broad based: i t decreased overall and inmost areas, and also the depth (poverty gap) fell by 28 percent, and the severity (squared poverty gap) fell by 34 percent. 100. Inrespect ofpoverty reductionperformance, Mozambique has overtaken some countries inthe region. To compare Mozambique's poverty rate with that o f its neighbors, we use the percent o f the population livingo n less than $1 (USD) inPPP terms, per day. According to this poverty line, 29 percent o f the population would be classified as poor in2002/3. Compared to its neighbors, Mozambique i s poorer than South Africa (7 percent in 1995), Tanzania (20 percent in 1993) and Uganda (25 percent in2002) butricher thanZambia (64percentin 1998),andMalawi (42 percent in 1998). Mozambique i s no longer the country with the highest poverty rates in the region. 101, Regionally, poverty reduction was greatest inthe Center, especially inrural areas.63 Poverty increased inthe South, es ecially Maputo, as well as the surrounding province and inurban areas inInhambaneqr4 The small poverty reduction inthe rural South was overwhelmed by the increase inurbanpoverty. 102. Poverty decreased unequivocally between 1996/7 and 2002/3, as emerges from the summary in Table 13. The mean percentile growth rate o f 4.1 percent i s the growth 63 Government analysis attributes the increase in Cab0 Delgado to poor sampling in both years but primarily inthe earlier survey, which led to an underestimationo f poverty in the 199617 data. 64This findingshouldbe tempered: ifthe Government team hadusedthe same food basket for both survey years, poverty inMaputo City would have fallen by 2 percentage points, rather than increased. But the change inpoverty nationwide would have been smaller (-6.2 compared with -1 5.3 with the new basket). The change inbaskets corrects for the large change inrelative prices which occurred between the two surveys, partly due to the devaluation, which raised the price o f imported food and non-food items. The Government paper argues that there was an increase in the quality o f food consumed inthe cities, which may have led to the finding o f higher poverty rates. 31 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM rate o f consumption of each percentile o f households, between 1996/7 and 2002/3, and then averaged over all percentiles. The Ravallion-Chen (2003) "rate o f pro-poor growth" i s the mean growth rate o f the poor, which i s also positive, although less than the growth rate at the mean or the median, reflecting rising inequality. Table 13. Growth of consumption and pro-poor growth Growth rate over the period Annual growth rate (%) 1996-97 to 2002-2003 (%) Growth rate in mean 30.9 4.6 Growth rate at median 27.4 4.1 Mean percentile growth rate 27.2 4.1 Rate of pro-poor growtha 25.7 3.9 Source: Fox el al. (2005). aAssumingaheadcount index of 69.1percent, and apoverty line of 11,240 Mt./day in2002 real terms. Non-income welfare trends 103. Other measures o f well-being confirm that poverty has declined. The share o f household expenditures on food has declined, indicating that poverty has fallen, because as households get richer they spend relatively more on non-food (Table 14). Households now own more durable goods such as radios and bicycles and have invested inhome improvements. Noteworthy i s the improvement inthe proportion o f houses which have been fitted with a better roof (16 percent up to 25 percent). Inthe poorest quintile, this rose from 11percent to 27 percent, a good indicator o f increases inwealth and welfare.65 Table 14. Non-monetary measures of welfare (percentage) All Urban Rural 1996 2002 2002 2002 Food share 68 61 50 66 Durable goods Radio 29 65 65 65 T V 5 9 23 1 Clock 24 43 60 34 Motorbike 1 2 3 1 Bicycle 13 40 23 50 Housing Durablewall 31 85 87 85 Durableroof 16 25 56 11 Source Fox et al (2005) See Annex Table 11, p 141, for a breakdown by quintile 104. Public policy has played an important role inimproving welfare. The use o f safe water (Le. private or public tap water and protected springs) increased from 24 to 37 See Annex Table 11, page 141 for the breakdown of Table 14 by quintile. 32 CHAPTER3: POVERTY percent, the distance to water sources decreased, and latrine usage increased (Table 15). School enrollments increased dramatically (as has already been pointed out indiscussing the role o f foreign aid, Table 10, p. 26).66 Table 15. Access to services p e r c e n t a g e 1996 2002 Water Use safe water 24 37 Less than 30 minutes to water 69 90 Sanitation Latrine 35 45 Education Enrolled 7-12 51 93 Enrolled 12-18 41 69 Source: Fox et al. (2005). See further detail, and abreakdownby consumptionquintile, inAnnex Table 12, p. 142. 105, Access to services has improved: attended births rose from 26 percent in 1993 to 48 percent in2003, and vaccination coverage rose from 55 percent in 1995 to 82 percent in2003 (Table 10, p. 26). Despitethe increased"health effort", health outcomes show a mixedpicture. Infant and under five mortality rates have improved in all quintiles (Table 16), and contraceptive use has risen (Table 10, p. 26). But total fertility rates (TFR)have barely moved as a strong decrease inurban areas was balancedby a slight increase inrural areas. HIV/AIDS prevalence has risen to 16.2 percent. Malnutrition continues to be serious, with 41 percent o f children stunted and 4 percent suffering wasting. The outcome o n malnutrition is not exceptional. Income growth alone will not be sufficient to meet the MillenniumDevelopment Goal o f halving the prevalence o f underweight children (low weight-for-age), and direct interventions will be necessary (Haddad, 2003). 66 See further detail, and a breakdown by consumption quintile, inAnnex Table 12, p. 142. 33 MOZAMBIQUE COLWTRY ECONOMICMEMORANDUM Table 16. Selected health outcome indicators Total fertility rate (TFR) Infant mortality Under 5 mortality 1997 2003 1997 2003 1997 2003 Total 5.6 5.5 147 124 219 178 Urban 5.1 4.4 101 95 150 143 Rural 5.8 6.1 160 135 237 192 Poorest quintile" 6.3 188 143 278 196 Poorer quintile 6.1 136 147 214 200 Middle quintile 6.3 144 128 216 203 Richer quintile 5.2 134 106 187 155 Richest quintile 3.8 95 71 145 108 Source: Fox et al. (2005); in tum from Demographicand HealthSurvey, 1997 and 2003; Gwatkin et ai., 2000. aQuintilesare wealth quintiles. TFR total fertility rate for ages 15-49, expressedper woman. Inequality 106. One key reason for the strong poverty performance can be seen inthe growth and distribution of consumption. Overall, consumption according to the household surveys grew at an average annual rate o f 4.6 percent, which i s slightly higher than the growth o f private consumption measured inthe national account^.^' Table 17 shows that inrural areas, the average consumption o f the bottom quintile grew less than the other quintiles: 21.6 percent for the period, comparedwith around 30 percent inall other quintiles. In urbanareas, real consumption growth was highest inthe bottom andtop quintiles and lower inthe middle three quintiles. As a result, growth did not translate into as much poverty reduction inurban areas. Table 17. Growth of consumptionby quintile, 1996/7-2002/3 (percentage) 1'' quintile 2ndquintile 3rdquintile 4'h quintile 5" quintile Totar- Rural 21.6 30.0 31.1 31.1 30.1 27.5 Urban 27.0 11.2 14.8 16.5 28.2 24.4 All 23.4 25.6 27.8 28.1 36.1 30.9 Source: Fox et al. (2005) 107. These changes inthe distribution o f consumption can be summarized by the measures o f inequality, the Gini and the Theil (Table 18). For Mozambique as a whole, there was a small increase ininequality over the period.68 The movement inthe Ginis was less than the movement inthe Thei1.69 Inequality inurban areas i s substantially 67 See the discussion about the national accounts figures for consumption, footnote 45, page 14. '*Thiscan also be seen inthe growth incidence curve, Annex Figure 14, p. 144. James and Amdt (2005) claim that the increase was too small to be statistically significant. 69 This may be because the Gini coefficient in effect weights more heavily changes around the middle o f the distribution, whereas the Theil stresses the lower end o fthe distribution more. 34 CHAPTER POVERTY 3: higher than withinrural areas. Despite the increase, Mozambique's overall level of inequality i s one o f the lowest inAfrica.70 Table 18. Measuresof inequality,1996/7 and 2002/3 199617 Theil Gini Urban Rural Source: Fox et al. (2005) 108. Are Mozambique's regions becoming more unequal? Concern has been expressed about a possible bias of growth infavor o f the major cities o f Maputo and Beira. As o f 2000, per capita GDP was much higherinthe provinces which include large cities (Sofala, at $193, and Maputo at $214, with Maputo City at $1,137).71 Elsewhere, since the economies are overwhelmingly rural usingsimilar technologies, the provincial GDP per capita within betweenthe narrow compass o f $85 and $137. Ifthere were a bias towards Maputo and Beira, this would contradict the standard finding inthe international literature that regions with lower GDP grow faster, conditional upon other growth-enhancing factors. Examining the hypothesis inMozambique turned up three main findings (see Benito-Spinetto and Moll, 2005): > Usingprovincial-level GDP data for 1996-2000, there i s no relationship between starting GDP per capita, byprovince, and subsequent Projecting forward, this would mean increasing inequality o f GDP per capita at the provinciallevel. 9 Usingprovincial-levelinvestment databetween 1996and200073,there isno convergence. Infact, Maputo province has a muchhigher investment rate per capita than the others.74 9 Usingprivate consumption data for 199617-200213,there isno convergencefor rural areas, but there i s some convergence for urban areas. 109. Overall it i s hard to conclude anything due to the data limitations. Accordingly no policy conclusions canbe drawn. Inany case, it should be noted that regional inequality may be driven by the process o f growth. I t maybe that the best resources -in terms of institutions, infrastructure, labor, management, etc. -were available inthe better-off areas, and for this reason the better-off areas attracted more investment. Hence there might be a welfare cost attaching to attempts to slow the divergence - ifdivergence it is. However, it is worth investigatingwhether through targeted infrastructural ' O See Annex Figure 15, p. 145, for a comparison o f Mozambique's inequality with other African countries. 71 See Annex Tables 4 and 5, p. 137. 72 See Annex Figure 6, page 138. 73 ConservativelyexcludingMOZAL. 74 See Annex Figure 7, p. 139. 35 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM investments, tax incentives, and other methods, ways could be found o f directing investments to poorer provinces without "killing the goose that lays the golden egg". c. ACCOUNTINGFOR THE CHANGEINPOVERTY 110. The major drivers o fpoverty reduction were (a) good results inthe populous Center and North, (b) a reduction inrural poverty, and (c) a reduction inpoverty o f agricultural households. This emerges from Table 19, which shows the contribution to total poverty reduction o f the growth inconsumption o f various groups, weighted by the share in the pop~lation.~'From the decomposition o f aggregate sectors, the change in poverty o f households whose head works inagriculture accounts for fully 11percentage points o f the national reduction inpoverty o f 15 percentage points. Poverty continues to be primarily a rural phenomenon because the bulk of the population i s rural. Table 19. Decomposing poverty changes (geographical and sectoral dimensions) Poverty in 1996 (%) 69 Aggregate sectors (%I Poverty in 2002 (W) 54 Change inagriculture poverty -11 Total change inpoverty 2002-1996 (%) -15 Change inindustry poverty -0.7 Change inservice1poverty -0.9 Regional decomposition ("/.I Changeinservice2poverty -0.8 Change inpoverty inthe North -3.4 Total intrasectoral component -14" Change inpoverty inthe Center -12 Change inpoverty inthe South 0.2 Urban-rural (%) Total intraregional component - l j a Change inurbanpoverty -2.5 Change inruralpoverty -13 Total intrasectoral component -1.5" Nole: Individuals are assignedto the sector where the householdheadis employed. `Service 1' includestrade, transportand services; `service2' includes health, education, and public administration. `The decomposition also allows for population shifts and an interactionfactor. As these are small they are not reported. Source: Annex Table 17 (D. 147L in tum from Fox et nl. (2005). Table 11. 111. The link between the agricultural growth noted in Chapter 1(page 17ff) and poverty reduction inrural areas emerges fiom a graph o f growth inmaize and pulse output versus poverty reduction (Figure 5). There appears to be a reasonable correlation between changes inpoverty and changes infood grain output, despite the expected problems of m e a s ~ r e m e n t . ~ ~ 7' See the full details in Annex Tables 16 (p. 146) and 17 (p. 147). 76 That is, ina perfect world all observationswould be inquadrants 2 and 4, but in fact there are some in the other quadrants also. 36 CHAPTER3: POVERTY on 0 0 10 - n Y 50 100 150 L 0 -" Growth in cereal and pulse production, 1996-2003(%) , Figure4. Foodcrop production and rural poverty,1996-2003'' Determinantsof income andpoverty, 2002/3 112. Poorer households canbe characterized by a number o f demographic and social factors (Table 20). They are more likely to be large. They have a larger number o f dependants per working adult. They are more lik.elyto be headedby a woman. They are more likely to have a disabled adult, and are characterized by low educational levels. The economic dependency ratio increased inall quintiles, especially inthe bottom quintilewhere it rose from 1.1 to 1.5. Education- at least low levels ofit, mostlyjust partially completed EP1-has become more common: inthe bottom quintile 50 percent o f household heads had no education at all in 1996`7, and this fell to 38 percent in 2002/3. While the poor are most likely to be employed in agriculture, this too i s changing: 87 percent of the bottom quintile were inagriculture in 1996/7 and this fell to 80 percent by 2002/3.78 The top quintile works in agriculture, trade, services, education and public admini~tration.'~ 77 Source: Jeeva Perumalpillai-Essex and Josef Loening (2005), in turn from staff estimates based on IAF 1996/7 and 200213 poverty data from Fox et al. (2005), and TIA 1996 and 2003 production data from Mather et al. (2005). The observations refer to provinces. Sofala province i s omitted as an outlier for both grains and pulses, and Manica province i s also omitted for pulses. This makes 9 observations (provinces) for grains and 8 for pulses. The estimated equation is: poverty-reduction = -11.9 - 0.175(crop-growth), with standard errors 2.9 and 0.039, and R2=0.57. "SeeAnnexTable18,p.142,fordetailsaboutsectorofemploymentbyconsumptionquintile. 79 See Annex Table 18 (p. 148) for details on employment by consumption quintile. 37 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Table 20. Householdcompositionby consumptionquintiles 1 ~ All 1'` quintile 5th quintile 1996 2002 1996 2002 1996 2002 Demographics Household size 4.8 4.8 6.4 6.0 3.5 4.0 Female head 21% 25% 20% 27% 23% 25% Disabled adult(s) 7% 7% 9% 10% 6% 5% Economic dependency ratio' 0.99 1.23 1.14 1.54 0.77 1.04 Head characteristics Head no educationb 41% 29% 50% 38% 32% 21% aEconomic dependency ratio, i.e. number of peoplenot workinginumber of people working. Source: Fox et nl. (2005), Table 13, which containsthe full table, including all the quintiles and several more variables. 113. The key determinants o f household consumption are education, demography, and employment sector, as emerges from a regression analysis.80 Education o f headhas the expected positive signs, with convex returns (viz. larger effects per year o f education as the level o f educationrises), reflecting the relative scarcity o f secondary and post- secondary education. The impact o f education i s higher inurban areas than inrural areas for all levels o f education. 114. Most demographic factors have effects that are roughly the same inrural and urban areas. The only exception i s the number o f menbetween 15 and 59 years old. This suggests that men inrural areas may not bring in as much interms o f household consumption as they take out, whereas men inurbanareas do. The presence o f disabled adults has a negative effect on household consumption inrural areas, as it adds to the dependency burden o f the household. Households with a widowed head have less consumption inurban areas. As far as sector of employment i s concerned, households in agriculture are among the lowest-consuming. Some o f the sectoral differentials are quite large - o f the order o f 20 to 30 percent. Inurban areas, working ineducation seems to gamer consumption equal to that working inagriculture, which seems to suggest an emerging teacher pay issue. 115. Structural changes include a decline inthe payoffto education inurban areas at the primary levels, and an increase for post-secondary education levels. The premiums for working intrade or services decreased inboth rural and urban areas; this is presumably associated with the large observed increase inemployment inthese sectors. The premiumfor working inthe manufacturing sector inurban areas fell from 16percent in 1996/7 to virtually nothing in2002/3; this may havebeen associated withthe privatizations and restructuring o f state enterprises which took place inthe late 1990s. Structureof income, livelihoodstrategies, and the labor market 116. Most households (88 percent inrural, 90 percent inurban areas) have at least one source o f cash income, and many have more than one, in addition to income received in kind(e.g. from subsistence agriculture or self employment). Although subsistence production i s by far the largest single source o f income inrural areas, non-agricultural self employment income, as well as sales o f agricultural produce and animal products The full consumption regressions are presentedinAnnex Table 19 (p. 149). 38 CHAPTER3: POVERTY provide an important share." Inurban areas, subsistence agricultureis much less important (only 12 percent o f total income), as employment income provides the largest share, followed by self-employment income. Pensions are a significant source o f income inurban areas, but they are virtually absent inrural areas. Remittances are a small source o f income (2 percent o f households). 117. Byquintiles there are some significant differences incash income sources. Not surprisingly, the top quintile gets less o f its income from subsistence agriculture (26 percent) than the other quintiles (40 to 48 percent). Employment accounts for between 20 and 30 percent o f income and there i s no obvious pattern across the quintiles. Self- employment provides 20 percent o f income inthe top quintiles, but only between 5 and 8 percent inthe other quintiles. 118. There have been dramatic inter-sectoralshifts inemployment. As would be expected inthe development process, the share o f agriculture, forestry and fisheries in total employment fell from 87 percent in 1996/7 to 80 percent in2002/3 (Table 21).'* This decline implies that the absolute number ofpeople active inagriculture grew only slightly between 1996/7 and 2002/3.83 The slack was taken up by trade, services and construction. Besides agriculture, the only other sector whose employment share declined was manufacturing: 2.3 percent to 0.8 percent. This was associated partly with the poor employment performance o f the formal sector, as we shall see below. Most o f the expansion of the labor force was absorbed by the informal sector, which grew at between 7 and 8 percent per annum. Table 21. Employmentby sector and by ruralhrban location, 1996/7 and 2002/3 1996/7 a 200213 a Rural Urban All Rural Urban All P e r c e n t a g e o f w o r k f o r c e Agric., forestry & fisheries 94.7 41.4 87.3 89.7 28.4 79.9 Mines 0.4 1.8 0.6 0.5 1.o 0.5 Manufacturing 1.2 9.2 2.3 0.5 2.3 0.8 Construction 0.5 4.6 1.1 1.3 6.6 2.1 Transport 0.2 4.3 0.8 0.5 4.1 1.o Trade 1.1 16.4 3.2 4.0 25.0 7.4 Services 0.8 11.8 2.3 1.9 22.6 5.2 Education 0.5 2.3 0.8 1.2 3.8 1.6 Health 0.2 1.9 0.5 0.3 1.4 0.5 Public administration 0.3 5.1 0.9 0.4 4.9 1.1 Total 100 100 100 100 100 100 Source. IAF 199617and 200213, Fox et al. (2005). The 199617definition o f "urban" was used inboth 199617and 200213. Urban areas were extendedby some 50 percent in the interveningperiod. Age group: 10-59years. aEmployment includes self-employment, family employment and employers. For abreakdownof this table by poor and non-poor, see Annex Table 23 (p. 153). Diagrams detailing the source o f income are given in Annex Figures 20 and 21 (p, 150). The latter gives these by quintile. 82 See also a table on employment by employer type (Annex Table 3, p. 137). 83 Using INE's figure o f population growth o f 2.4 percent per annum (Table 1, p. 2), this would imply growth of the agricultural labor force o f 0.98 percent per annum between 199617 and 200213. 39 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM 119. Poverty declined the most for households whose heads were ineducation (53 percent to 27 percent), health (61 percent to 43 percent), and transport (54 percent to 35 per~ent),'~The increase inreal public sector wages probably explains the decline in poverty ineducation and health sectors. 120. Formal sector employment growth -not unexpectedly -was very low indeed. Panel data surveys o f the manufacturing sector give a growth elasticity o f employment o f only 0.1,85 which i s below the average in other African and developing countries.86 What this means, inplain terms, i s that ifthe economy grows by 10 percent, formal sector employment grows by only 1percentg7 This was in a context o f swift growth o f firm sales - o f the order of between 6.5 and 30 percent per year. For lack o f data, i t i s not known whether formal employment in other sectors (e.g. services) grew at faster rates. 121. Does labor law hamper employment in theformal sector? Mozambique has the eighth least flexible labor market in a survey o f 145 countries (World Bank 2003e).** In almost every dimension, Mozambique's regulations are more restrictive than the region's, particularly inrespect o f severance pay and dismissal noticeeg9Retrenchment i s costly and time-consuming: it costs 141 weeks o f wages to retrench a worker in Mozambique, while the average inAfrica i s 60 weeks' wages." To the extent that this cost i s binding, making dismissals expensive may have the unintended effect o f discouraging hiringin the first place. Almost 45 percent o f foreign-owned firms claimed that they were being forced to retain excess workers on account o f the highcost o f retrenchment (World Bank, 2003a, p. 35) There are further problems. Piece rates, a standard way o f linkingpay to productivity, are not allowed. Hiringexpatriate employees, despite the changes introduced by Decree 57 o f 2003, i s cumbersome and expensive. 122. There would seem to be aprimafacie case for bringing Mozambican labor law into line with practice in the region, because: P Mozambique is an outlier interms of low labor absorption-hiringbythe formal sector was lower than incomparable countries - despite highinvestment rates and vigorous value added growth; 84 See Annex Table 22 (p. 152) for poverty rates by sector of employment o f the householdhead. Also see a discussion o f the decline o f poverty inthe mining sector, Table 37 (p. 96). 85 The aggregate statistics give slightly higher elasticities, but these are contaminated because they include an unknown amount o f informal sector employment also. For the record, private sector value added, excluding the megaprojects (but including an unknownamount o f informal sector activity), grew at about 6 percent per annum between 1996 and 2002, the growth of urban private sector employment (including an unknown amount of informal sector activity) was between 1 percent and 2.6 percent per annum. 86 See Benito-Spinetto and M o l l (2005) for details about other countries' elasticities o f employment with respect to growth. 87 It should be noted that these surveys were o f formal-sector firms that were unaffected by the restructuring o f state enterprises. ** See the full listing inDoing Business in 2004 at http:/.'im worldbank.orgiDoineBusinessiExploreTopics/HirinpFiringWorkers/ConlpareAll,aspx *'See Appendix Table 29 (p. 156) for the details o f several countries inthe region. 90 Mozambique is the world record-holder inthe amount o f severance given per year o f service (namely 1% months). See table Severance Payment inBetcherman et al. (2001, p. 31). 40 CHAPTER3: POVERTY > Mozambique i s an outlier inthe restrictiveness o f its labor codes 123. The case i s not complete, however. There are gaps inour knowledge: P Inother countries the causation between labor market flexibility and employment growth i s not always clear. The standard wisdom o f the 1990s was that labor flexibility and employment growth inOECD countries" are causally related, but some recent evidence i s less than ~ompelling.'~Stronger evidence o f the negative effect ofjob security rules on employment comes from Latin America93where such rules have been tighter, making for measurable impacts. Across African countries, managers' perceptions o f labor as a significant problem are related, if weakly, to firing costs as defined in law; but they do not seem to be related to the "difficulty o f firing inde~"'~;the evidence for Africa i s too scant to permit generali~ation.'~This agnostic view does not mean that just any degree o f labor rigidity is acceptable; it means that a country-specific investigation is neededto establishthe case for increasing employment by reducing labor market rigidities. > Evidence i s lacking as to whether restrictive labor regulation i s a binding constraint inMozambique. Some 38 percent o f manufacturing business owners complained that labor regulations were a "major" or "very severe" problem (World Bank, 2003a), but other items were more serious - financing costs (84 percent), electricity (64 percent), corruption (64 percent), crime (54 percent), customs (49 percent) and so Inaddition, there is as yet no objective measure of the costs involved. By implication, it i s not yet possible to estimate the size o f the likely increase inemployment should the administrative barriers be lifted. Also, while more labor market flexibility may be a necessary condition for increasing employment, it may not be a sufficient condition. P There is a lack o f evidence about potential losers from the process. Even if greater flexibility helped raise employment inthe long term, it i s possible that in ''See Layard et al. (1994), OECD (1994),Nickel1 (1997), OECD (1999), Riboud et al. (2002). 92 See Baker et al. (2002). 93 Guasch (1999), Heckman and PagCs (2000), Saavedra and Torero (2000), Gill et al. (2001), World Bank (2003e) and Kugler (2004). Betcherman et al. (2001, p. 7)provides an excellent review. But see Downes et al. (2000) who find no impact o f rigidities on employment inthe Caribbean. 94 See Annex Figures 30 and 31 (p. 157)which give the detail and sources. 95 Fallon and Lucas (1991) find that a 10 percent increase in dismissal costs inZimbabwe increases long- termunemployment by 20 percent. N o further empirical studies of the subject inpoor African countries could be located. According to the WED surveys, business people inAfrica consider poor transport, red tape and power outages to be more serious than labor rigidities. As yet, there i s little evidence that labor legislation reduces formal sector job creation (Fox et al., 2004). Also using the WED, Alby et al. (2005) findno correlationbetweenlabor market rigidity as defined inlaw and firms' perceptions o f such rigidity as a problem. Eifert et al. (2005) find that Africa i s a high cost region - due to poor roads, corruption, unreliable electricity, etc.-with price levels 30 percent above levels predicted on the basis on income per head. See also the congruent findings by Lindauer (1999) that regulatory reforms are likely to make only a modest diference injob creation; by Forteza and Rama (2002) that mandatory benefits did not affect growth in over 100 countries; and by E C A Labor Market Team (2004, section 5.2, p. 59) that job creation inEurope andCentralAsia was affected by administrativebarriers andhightaxes -not labor market institutions. "Seeausefulsummaryofthedataathttp:!/rru.worldbank.orp`InvestmentClimate/ , 41 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM the short andmediumterm some employees would be let go. What would be the size of the trade-off, and what mitigating policies could be adopted inthe interim? 124. Inorder to investigate the likely outcomes ofreform, aPoverty and Social Impact Analysis i s being donejointly by the Government and the World Bank. Meanwhile, the Government i s preparing a modernization o f the labor law which will address the call for increased labor market flexibility. 125. There i s a bright side to the growth-employment nexus. Within the manufacturing sector, the panel surveys showed that newly created firms with an export orientation tended to create more employment than older firms or firms producing for the local market. This may indicate that older firms are still restructuring and shedding workers. It could also suggest that the set o f policies followed - a realistic exchange rate, reduction o f tariff barriers, and other investment climate factors - have helped enhance job creation. 126. As mentioned above, the employment share o f manufacturinghas declined. In absolute numbers, employment inthe sector was some 120,000 in 1996/7, falling to some 50,000 in2002/3. By employer type, 55 percent o f these (66,000) were employed inthe private sector, the public sector, or government in 1996/7, and 62 percent (31,000) in 2002/3. The latter category can be taken as a proxy for the formal sector portion o f manufacturing. The decline o f more than half arose from two sources: (a) the closure and privatization o f erstwhile state enterprises inthe mid- and late 1990s resulted injob losses, and (b) among the surviving- and growing - firms, employment growth was low. 127. Employment shifts are consistent with theoretical predictions, as noted above. Table 22 shows average annual employment growth rates, value added growth rates and value added per unit o f labor growth rates by four sectors, i.e. agriculture (including fishery), industry, private services (transport, trade, services) and public services (education, health, public administration). This table shows that labor has shifted out o f the slower-growing agricultural sector into the faster-growing service sector. Not surprisingly, aggregate labor productivity improved inagriculture and fell inthe sectors where employment expanded. It also improved inindustryas a result of the restructuring. Table 22. Average annualgrowth rates, employment and value added, by sector, 199617 to 200213 Avg. annual growthof Avg. annualgrowth of Avg. annual growth of employment GDP at factor cost GDP/ employment Agriculture -1.15 % 5.11 % 6.33 % Industry -1.10 % 16.43 Yo 17.72 % Services (private) 16.47 % 5.69 % -9.26 % Services (public) 8.81 % 7.41 % -1.29 % Source: Fox et al. (2005). Employment data from the IAFs. GDP from INE. 128. Wages inthe private formal sector increased only slightly, by about 1.6 percent a year inreal terms between 2000 and 2003, less than consumption growth o f 2.7 percent for the same period, and well short o f GDP growth o f 7.3 percent. 42 CHAPTER3: POVERTY 129. Expected developments inthe urban labor market inthe next ten years give little room for cheer. The working age population (ages 15-64) will increase by at least 2.4 percent per annum between 2005 and 2015 (Table 1, p. 2). With rural-urban migration, the urban working age populationrises at a higherrate, possibly as much as 4.1 percent per annum. Meanwhile, at best formal sector wage employment may increase at 1.5 per annum. The substantial gap between growth of the urban working population and employment implies boththat the informal sector will continue to expand and that formal sector wage levels (other than inpublic employment) are likely to stagnate. 130. There areprominent gender differences in employment. The majority o f those in the labor force are self-employed or working ina family business. Men are more likely to be wage earners or self employed, and women are more likely to be working ina family business, particularly in agriculture, transport, trade and services.97 Women are found more often inagriculture, meninthe traditionally male-dominated sectors o f manufacturing, construction, transport and public admini~tration.~~Both men and women are found intrade. 131, Factors that influence wage levels include education, marital status, and gender. This emerges from a regression analysis o fthe 16 percent o fworkers who are waged.99 The labor market penalty for beingfemale i s 28 percent. When women's andmen's wages are analyzed separately, the structure i s different but only the coefficients on education, contract type and the construction sector dummy were significantly different by gender. Marriage confers a premium for menbut not for women. Returnsto education also differ between men and women after primary school, but inthis case, women's returns are higher. As inthe consumption regressions, the payoffs to education are convex, but given the low numbers who are able to complete any kindo f post- secondary education, it i s possible that these coefficients simply reflect the selectivity bias of those who are able to go to schools rather than any effect o f educationper se. Being a casual laborer rather than having a regularjob seems muchmore disadvantageous for women. 132. Rural self-employment and rural wage employment provide roads out of poverty. Further regression work has been done usingagricultural household data sets (the TIA o f 2002) to examine the determinants of self-employment inrural areas, and the determinants o f rural labor market access and wage levels. The econometric work i s listed inPerumalpillai-Essex and Loening (2005, Tables 19 and 20) and neednot be repeated here. The key conclusions are these: (a) Factors increasing the probability o f self-employment include having a bicycle and the presence o f all-weather road access. Low education i s not a barrier to self-employment, but females are significantly disadvantaged. Once having entered into the labor market, landownership, access to credit and education are important determinants o f self-employment income. Education inthe rural self- employment sector has a rate of retumo f about 10.8 percent. 97 See the detailedbreakdowns o f employment by type of contract and gender, in Annex Tables 25 (p. 154) to 27 (p. 155). 98 Detaileddata on employmentby sector andby gender are given in Annex Table 24 (p. 153). 99 See the detailedregressionsin Annex Table 32 (p. 158). 43 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM (b) Education, infrastructure (roads) and market access1ooincreases the probability for generating rural wage income. Once having entered into the labor market, education and market access become key factors in explaining rural eamings. Educationhas here a rate o f returno f 14.8 percent. Insum, rural self-employment andruralwage employment appear to offer aroute out of poverty. However, there i s evidence that higher income non-farm jobs go to those with higher levels o f education and that woman are discriminated inthe non-farm labor market. An important component o f a comprehensive strategy for the development o f rural areas and alleviation o f rural poverty will be to deepen and extend the relationship with non-agricultural activities within rural areas. Non-farm employment income can expand only ifthe regional economies are diversified into year-round activities, including agro-processing, services, and manufacturing. A major limitation on such diversification i s the lack o f adequate infrastructure. D. CONCLUSIONS 133. Poverty declined, measured by both monetary and non-monetary measures. Progress inreducing monetarypoverty was substantial, as reflected inthe headcount as well as measures o f depth and severity, but it showed regionally unevenness. The most progress was recorded inthe Center, with less progress inthe North. 134. Non-monetary measures also show improvement - ownership o f durables jumped, the share o f food intotal expenditures fell for all quintiles, and many households living below the poverty line were able to improve their houses. Access to a safe water source and a latrine or other sanitation facility improved. School enrollments are up as well in all quintiles, although the poorest still lag. With respect to distance to services (health and education) a gap exists betweenthe top quintile and the other quintiles, except for distance to primary schools, which appears to be equal inall quintiles. So there i s scope to improve access to services for the lower quintiles. 135. One key reason for the good poverty performance i s that inequality did not change much, so aggregate growth inconsumption reached poor households. Overall, household consumption per adult equivalent grew by 4.6 percent per annum per adult equivalent, a very healthy rate, and the lowest quintile inrural areas seems to have had the fastest growth. Thus the modest increase ininequality only reduced the national poverty perfonnance by 8 percent. Inall regions, the strong growth inconsumption registered by agricultural households drives most o f the poverty performance. However, the shift o f households out o f agriculture as the main source o f income into other sectors also helps. 136. The poor inMozambique are mostly rural, with larger households and more dependents. They are slightly more likely to report that an adult member o f the household i s disabled. Overall, more heads are reporting that they are divorced and/or are female in the 2002/3 data, but this feature does not seem to correlate with monetary welfare. Education o f head does rise monotonically with consumption per capita, but even more noteworthy i s the sharp rise ineducation o f heads inall quintiles. Interms o f loo relevantvariableisadummyfor"marketinthevillage". The 44 CHAPTER3: POVERTY sector o f activity, many heads o f poor households have diversified into trading and other private services. 137. Multivariatetechniques indicate that the most important determinant o f household consumption per capita i s education o f head. This i s more important inurban than rural areas, but even inrural areas, education has a highpositive marginal effect on consumption. Controlling for education, working intransport, trade, or health increases household consumption o n the marginby about 30 percent. Inurban areas, working as a teacher decreases consumption relative to nearly all other sectors, suggesting that there may be a problem with teacher salaries inurban areas. 138. Household demographic characteristics has a lesser effect on consumption, and for the most part the estimated negative effects o f children and disabled adults were similar inrural andurban areas. Noteworthy i s the difference that inurban areas, widowhood has a strong negative effect for both male and female headed households. 139. Mozambicans' livelihood strategy i s to have a portfolio o f household activities. Most households have at least two earners with agriculture as a primary or secondary activity, and with agriculture usually done by women (90 percent work inagriculture), and often by men as well. Inrural areas, subsistence agriculture provides about half o f total income, with the rest coming primarily from sales o f agricultural products and employment income. Inurban areas, the majority o f household income comes from employment, followed by self employment. The fastest growing sectors o f employment were trade and private services, as labor moved out o f agriculture and into these areas. As a result, the absolute size o f the agricultural labor force fell, and labor productivity increased. This probably contributed to the strong poverty reduction performance inrural areas, Menare over three times as likely as women to be inwage employment, and 1.5 times as likely to be in self-employment. Only 16 percent o f the total labor force receives a wage, either as a casual worker or a salaried employee. Most women work ineither subsistence agriculture for home consumption, or in a family business. 140. Women inwage employment earn on average 28 percent less than their male counterparts. At higher levels o f education, women are actually rewarded better than men, but men and women are rewarded differently by sector - men earn more in agriculture, mining and transport. It i s not clear to what extent these differences reflect discrimination or unobservable differences injob content (e.g. mining), hours worked, or quality o f effort. 45 CHAPTER4. PRIVATE SECTORDEVELOPMENT: PRIORITIZATIONAND SEQUENCING A. INTRODUCTION 141. As remarkedinChapter 2, Mozambique hasbeen successful at attracting foreign direct investment inlarge scale projects known as "mega-projects" to exploit its mineral and energy resources. This has helpedfuel remarkably highreal GDP growth rates. But since the employment impact o f these investments i s limited,private sector-led growth in Mozambique will call for the broadening o f the base for growth, particularly through more rapid development o f small and mediumenterprises. 142. A great deal of valuable researchhas already gone into this important subject."' Gargantuan action matrices have beenassembled. The contribution o f this Memorandum i s not to rework this terrain but to amve at a workable synthesis and prioritizationo f the many disparate recommendations arising from these efforts. Thus this chapter is not the all-encompassing review o f all private sector- and business-related policies found in many Country Economic Memoranda; insteadit focuses on a small number o f key administrative barriers to growth. First, a brief description o f the private sector in Mozambique i s given. Then the criteria for selection o f priority policy areas are discussed, and then the eight items selected are described indetail. Inaddition, Box 1(p. 5 1) summarizes the main findings of the contemporaneous Growth and competitiveness study (World Bank, 2005c), so that these are not repeated inthis chapter. B. THEPRIVATESECTOR:A DESCRIPTION 143. The largest body o fprivate sector operators, and by far, are the three million-odd smallholder households. Muchthat i s recommended inthis Chapter i s likely to benefit this group, particularly the recommendationsontrade facilitation (see below, p. 51)as this would likelyreduce the transactioncosts for export andresult inhigher farmgate prices, To avoid repetition, however, the reader i s referred here to the substantial parts on agriculture (Chapter 2, p. 17 and Chapter 6, p. 110) and on rural land(Chapter 5, p. 72). 144. The second-largest body o fprivate sector operators comprises the informal sector inthe towns. This canbeinferredfrom the categories "self-employment", "family labor" and"employer" inAnnex Table 3 (p. 137) which sumto 60 percent o furban employment in 1996/7, risingto 76 percent in2002/3. The implied growthofinformalsector employment was between 7 and 8 percent depending on the assumption about urban population growth. (Remarkably, "employers" increased from 1.O percent o f the urban Some recent examples include: (i) Investment Climate Assessment (World Bank, 2003a); (ii) Doing The Business in 2004: Understanding Regulation (The World Bank and IFC, October 2003); (iii) Biggs and Fisman(1999); (iv) da Silva et al. (2002), (v) FIAS (2001); (vi) Nathan Associates (2002); (vii) Booz Allen Hamilton (2002); (viii) Mwangi and De Wulf (2003), and (ix) Matrices to the SeventhPrivate Sector Conference (Mozambique, March2003). MOZAMBIQUE COUNTRYECONOMICMEMORANDUM labor force in 1996/7 to 10 percent in2002/3, reflecting the growth o f small businesses, probably most of it informal.) 145. Our knowledge o f the informal sector i s extremely limiteddue to the absence o f specialized surveys. Data difficulties obstruct the drawing o f inferences about changes in earninglevels inthe informal sector.'02As the sector i s going to play an increasingly important role inthe future it will be essential, for policy-making purposes inthe fields o f education, health, small business promotion and poverty reduction, to conduct regular surveys of informal sector businesses, particularly inthe urban areas. 146. The third-largest subset of the private sector is the so-called formal sector. Most o f the ensuing discussion i s focused on this sector. As has been seen inChapter 2 (p. 14), the biggest sectoral shares inGDP are those o f commerce (22 percent), followed by agriculture (19 percent), manufacturing, construction and transport & communications. Growth o f value added was fairly strong inmost sectors - 5 percent per annum in commerce, 5.5 inagriculture, 18 (10) inmanufacturing including (excluding) the mega- projects, and 9 percent intransport & communications. As has beendiscussed, the biggest contributor to GDP growth was agriculture, followed by mega-projects in manufacturing, and then commerce, construction and transport & communications. In 2002 there were 31,735 private sector firmslo3 employing 310,000 people (NE 2002). O f these 396 firms (1.4 percent) were "large", employing over 100people, 9.1 percent were "medium" (10-99 employees) andthe rest were "small" (1-9 employees). 147. Mozambique has established itselfas one of the leading recipients o f direct foreign investment (DFI) inthe region. The main sources o f DFIwere South Africa (about 28 percent), the UK (22 percent), and Portugal (19 percent), followed by Japan, the United States and Mauritius (United Nations, 2001). The key sector involved was manufacturing, followed by agriculture/forestry/fisheriesand tourism. The majority o f investments inmanufacturing were directed at the local market, mainly infood & beverages and construction. DFIincreased sharply inthe late 1990s as "mega-projects" - initially aluminium processing, later mining-were set up; this was discussed inChapter 2 (p. 23). 148. A weakness that was remarkedon above i s the poor export capability of Mozambican firms. Although exports rose sharply after 1998, most o f this was due to the mega-projects; the export performance o f the remaining firms has beenweak, and only 7 percent of all output o f manufacturing firms i s exported (World Bank, 2003a). Moreover, the weak local private sector has not beenable to take advantage o f opportunities inthe domestic market. Owing to unfavorable investment climate conditions and economic inefficiencies, the range and quality o f services andproducts remains narrow, while prices tend to be highin comparison to Mozambique's neighbors. Not surprisingly, Mozambican consumers purchase a substantial amount o f goods and services inthe South African town o f Nelspruitnear the Mozambican border. 149. The Investment Climate Assessment (World Bank, 2003a) and the contemporaneous work on value chain analysis (World Bank, 2005c) show that this lack o f competitiveness i s partly due to investment climate constraints. Inparticular, the cost `02 See Benito and Moll (2005, para., 56, p.22) for a discussiono f this issue. lU3 non-profitsandgovernment. Excluding 48 CHAPTER 4: PRJORITIES FORPRIVATE SECTORDEVELOPMENT and quality o f infrastructure services, including electricity and transportation, impact directly on manufacturing costs o f production, and the highcost o f capital, coupled with highcollateral requirements, makes it costly to produce goods and services. 150. H o w can the reforms proposedinthis chapter be expected to impact on poverty? A disclaimer must first be given. The objective of this chapter was not to investigatethe impact on poverty of private sector development but to arrive at an ordering o f actions to facilitate development. To evaluate empirically the impact o f these actions on poverty would require unusually rich data sets and a large computable general equilibrium model. Nevertheless some pointers can be given here. 151. First, it will bemany years before the expansion of formal private sector employment will have a direct impact on poverty, that is, by the formal sector's drawing inlaborers fromhouseholds previously classifiedas "poor". This is simply because the formal sector i s as yet a small fraction o f overall employment, and the workers first in line are likely to be better educated, urban, and experienced. This can be seen inTable 23 which shows employment status o f the labor force, categorizedby the quintile o f consumption o f the worker concerned. The better-waged formal sector comprises, mainly, people inthe top two quartiles in urban areas, who numbered 11percent o f the quintiles in 1996/7,increasing to 12-13 percent in2002/3. (People below the third quintile are below the poverty line, and their wage levels are consistent with informal sector activities.) The point i s that as the formal sector grows, it i s likely to draw labor not from the lowest quintiles o f self-employed and non-private waged work but from the upper two quintiles, which workers are already above the consumption-based poverty line. Table 23. Employment status by consumptionquintileand location, 199617 and 200213 Urban Rural 1996!7 employment status: p e r c e n t a g e p e r c e n t a g e Self-employed 86 I 9 75 69 63 96 96 96 86 95 Wage non-private 10 13 14 20 26 2 2 2 2 3 Wage private sector 4 8 11 11 11 2 2 2 2 2 Total 100 100 100 100 100 100 100 100 100 100 200213 employment status: Self-employed 76 73 74 68 56 95 9 1 91 96 93 Wage non-private 16 18 16 20 31 2 2 2 3 5 Wage private sector 1 8 9 10 12 1 3 1 3 1 1 1 2 Total I 100 100 100 100 100 I 100 100 100 100 100 ISource: Computationsunderlying Fox el nl. (2005), usingmicrodataof the householdsurveys. Weighted averages; for all individuals betweenage 15 and 59 Q1 is the poorest consumptionquintile, and Q5 the nchest 152. Formal sector expansion will, in addition, have indirect impacts and inthe short term these are more important. As workers are drawn into the formal sector from the upper quintiles of self-employed and non-private occupations, they will inturnbe replaced by workers from lower quintiles; this i s nothing other than the standard "ripple effect". Furthermore, as more workers earn better wages inthe formal sector, they and 49 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM the firms they work for will demand more services, some o f which are drawn fiom the self-employed and informal sectors - food andbeverages, transport, domestic service, and the like. 153. Some specific pointers may also be given about the poverty impact o f the measures contemplated below. Trade facilitation i s intended to assist inexportation o f all kinds, and as noted above, would probably leadto better farmgate prices, andmore employment inlabor-intensive industry. The other measures, and inparticular improving access to energy and improving access to finance, are likely to improve the health o f the private sector ingeneral, and hence result inthe indirect effects on poverty identified above: the ripple effect o f employment and increases indemand for services. c. CRITERIA FORESTABLISHINGTHE PRTORITIZATION 154. There i s a need to focus the scarce resources o f the private and public sectors o n removingbarriers that are bindingand can be relatively easily removed, and will contribute to time and cost savings indoing business inMozambique, while enhancing the attractiveness o f Mozambique as an investment destination for small, medium, and large scale investments. One guide to what i s "binding" i s given by manufacturing firms' opinions (Table 24). The cost offinancing is considered the biggest barrier, and while this measure is hardly objective -what businesspersonwouldnot prefer cheap money? - the measure is a meaningful one inthe "investment climate" context, for inother countries with no bigbankingsector problems, this measure i s rated low (the cross- country average i s 38 percent). Hence this factor i s included inthe "short list". Table 24. Firms'perceptionsof "major" or "severe" obstacles to business,2001 -- o e r c e n t a ~ eo f firms rating the obstacle as " m a j o r " or "severe" -- u J 1 1 Cost of financing 84 Tax rates 55 Business licences and 28 operating permits Electricity 64 Crime, theft and 54 Transportation 27 disorder Corruption 64 Customs and trade 49 Access to land 27 regulations Macroeconomic 63 Tax administration 47 Telecommunications 21 instability Anti-competitive or 60 Labor regulations 38 informal practices Economic and regulatory 58 Skills and educationof 34 policy uncertainty available workers Source: World Bank (2003a), summarized at littn:/~nu.worldbank.o~g/lnvestmet~tC:l~~ 155. Electricity was rated at 64 percent and so i s discussed below. Corruption (64 percent) i s next. This, together with the other governance measures - "anti-competitive or informalpractices" (60 percent), and "crime, theft and disorder" (54 percent) - are not treated in general inthis Memorandum as they are specialized topics, but one important component, namely the difficulty o f contract enforcement, i s discussed below. Macroeconomic instability (63 percent) has been discussed inChapter 2. Tax rates (55 50 CHAPTER 4: PRIORITIES FORPRIVATE SECTOR DEVELOPMENT percent) are discountedbecauseinfact Mozambique's tax rates are, on the whole, (too) low, as discussed inChapter 2. Customs and trade regulations (49 percent) are discussed below under the rubric o f "trade facilitation". So the four items pinpointed -the cost o f finance, electricity, trade facilitation, and contract enforcement - form the core o f the prioritization. 156. Inaddition, there are some "quick wins" that canbemade. These will not necessarilymake the biggest impact on investment or growth, but they can be achieved fairly quickly and at low cost, and would help to build goodwill between the government andthe business community. They are: accelerating businessregistration, adopting a new Commercial Code, accelerating value added tax refunds and reducingthe cost o f inspections and fines while improving compliance. A sequencing of actions to achieve these aims is presentedinTable 28, p. 68. Box 1. SUMMARY OF GROWTH COMPETITIVENESS AND This paper (World Bank, 2005c) uses detailed value chain analyses o f cotton-to-garments industry,wood processingandtropical h i t s (banana andmango). Inthe cotton-and-garments industry, the two mainissues that impactedthe entire supply chain are the cost of transport and the low labor productivity andtechnical skills. The cost o f transport dominates many portions o f the value chain as the lack o f a north-south transport corridor forces altemative routing and results inhigher transport costs and longer durations. A north-south transportation corridor needs to be developed. Labor productivity inthe production o f T-shirts i s 10-11items per day, compared with Kenya's 20-24. Targeted subsidies for an ambitious program o f skills upgrading could help stimulate investment. Public-private partnerships inskills development may be a useful vehic1e. Inthe wood-processingchain, cumbersome steps and fees plague the export process which makes up 27 percent o f value added between forest and mill. Barriers (such as the lack o f organized information on forest resources) could be reducedby computer-based tracking for forest resources, a transparent management system for the concession approval process, and clearly written guidelines for potential concessionaires. Inthe tropical fruits value chain, the mainrecommendationare that all stakeholders should club together to organize cable systems (bananas) and to set up a regional agrochemicalbuyingstrategy for cheaper, bulk prices. D. IMPROVING ACCESS TO FINANCEAND REDUCINGITS COST 157. The cost o f finance is the most serious constraint to the development o fthe Mozambican private sector. Interest rate spreads are high(reaching 19 percentage points in2002, falling to 12-15 percent in2005), andrealinterest rates are highandvolatile (16 percent in2002, 15.5 percent in2005). As o f the time o f the Investment Climate Assessment surveys, Mozambique's real interest rates were higher than those o f countries 51 MOZAMBIQUE COLWTRY ECONOMIC MEMORANDUM' at similar levels o f development which averaged 9.4 percent (Figure 5).'04 A decomposition o fbank spreads reveals that 44 percent o f Mozambique's bank spreads are explained by loan-loss provisions and 35 percent by highoverhead costs. 16 - 0 Hondu E h 1 4 - Tanzania v) 2 0 1 2 - Nepal @Cambodia Pakistan C Bolivia 0 -2 Mozambique0 Nigeria Bhutan 0 1 0 - 0 a, E Uganda 0 %angladesh 5 0 China P F a - 0 Eritrea .- ; 0 Zambia 8 6 - 0 Kenya 0 100 200 300 400 500 600 700 800 900 1000 GNI per capita in 2001 (US$). Figure5. Real interest ratespaidby manufacturingfirms - intercountry compari~on'~~ 158. Giventhe highcost of finance, most Mozambican enterprises rely on their own resources for investment and working capital. Only 29 percent o f firms have bank loans. Inthis regardthere isplentyo froom for improvement; there are several countries at about Mozambique's level o f er capita income with much better access to bank loans, e.g. Eritrea with 45 percent."' Only 12percent o f Mozambicanmanufacturing firms have overdraft facilities, less than the nine African countries surveyed whose average was 43 percent.'07 InMozambique, some 13 percent o f new investment i s bank-financed, less than the Africa average o f 21 percent. Smaller firms face greater barriers in obtaining external financing due to hightransaction costs related to problems o f information, communications and enforcement. Credit i s o f short duration, on average 35 months, insufficient for new investment. Collateral requirements are stringent, averaging 141 percent o f the loan amount, which is higher than inmost countries at a similar level o f development.lo' It i s small wonder that, as shown inthe Investment Climate lo4 source:InvestmentClimateAssessments, athttp:/irm.worldbank.orcr/InvestmentClimate/.Note Data that Mozambique's GNI and GDP per capita have increased since the time o f these surveys. lo5 InvestmentClimateAssessmentsandassociatedExceltables, at Sources: http://rru.worldbank.or~/InvestmentClimate/,The real interest rate is the average o f the nominalrates paid by the full sample of ICA survey firms, less the consumer inflationrate for the relevant year (between 1999 and 2003). lo6agraphofloanaccessbymanufacturingfirms, bygrossnationalincomepercapita:AnnexFigure See 34 (p. 160). I O 7 See ht~:~irru.worldbank.orcriInvestnieiitClimate,'Ex~loreTopics~Fii~aiice,aspx. lo* AnnexFigure36(p.161)foranintercountrycomparisonofcollateralrequired. See 52 CHAPTER 4: PRIORITIESFORPRIVATE SECTOR DEVELOPMENT Assessment (World Bank 2003a), the lack o f access to and the highcost o f finance was cited by 78 percent o f firms as a severe problemto their operations and growth. 159. Recommendations. Inorder to remedythis situation and stimulate the commercial banks inmakingsound new credits inthe difficult Mozambican lending environment, immediate and medium-term policy reforms will needto be adopted. Based on the recommendations o f the Financial Sector Assessment (FSAP) findings, the main areas o f focus are as follows. 160. Strengthening banking supervision. Some progress has taken place inthis area over the past two years, but muchwork remains. As most Mozambican banks are foreign-owned, the BankingSupervisionDepartment needs to establish regular communicationwith the head offices o f the foreign banks, and the bankingsupervisors in their countries ofdomicile. The current loan classificationand loanloss provisioning systems need to bebrought in line with international practices. 161. Enhancing the scope ofjhancial intermediation. Inorder to stimulate the commercial banks inmaking sound new credits inthe difficult Mozambican lending environment, several short- and long-term actions would needto be undertaken. P The establishment o fa commercialcourt, dealing with the largest cases. k Simplification ofjudicialprocedure. Thisisparticularly important, giventhe lack o f a trainedjudiciary. k The enhancement ofthe scopeandreliabilityofthe credit registryadministered by the Bank o fMozambique. P The reform ofthe landregistry and the establishment of aregistry for movable property. 162. Monetary and debt management. Although the sharp tightening o f monetary policy that took place inmid-2001 appears to have been effective instabilizing prices and the exchange rate, it contributedto the highlevel and volatility o finterest rates mentioned above, Inorder to reduce volatility o f prices and interest rates and enhance the scope for local currency intermediation- thereby limiting financial dollarization - the Banco de Moqambique needs to pursuemore transparent, pro-active (yet smoother), forward-looking, monetary management. 163, To achieve this goal, the current monetary framework needs strengthening. Market participants apparently do not fully understandthe goals or procedures o f monetary policy. This partly reflects insufficient efforts by the Banco de MoGambique (BM)to communicate and explain its goals and operating procedures, as well as flaws in the designand operation ofits monetary instrumentsthat sendconflictingsignals on the Government's stance on monetary policy. Acting preemptively to limit deviations from monetary policy objectives will also require a strengtheningo f BM's analytical capacity and a greater emphasis on inflationtargets and developments. At the same time, it i s important to rely more closely on an intermediate monetary target. To enhance transparency and avoid interfering with market signals, BMmay need to conduct its not particularlylow given its level of income-- see Annex Figure 34 (p. 158). `09 Mozambique's level o f creditrelativeto GDP, at 18 percent, has muchroom for improvement, but i s 53 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM monetary operations inthe overnight money market.' loAdditionally, adequate care needs to be exercised to reduce the significant crowding out o f private borrowing by Government borrowing. 164. The capital market inMozambique i s small and offers a limited scope for development in the short term. At this stage, the first objective should be to develop the market for public securities, thereby providing a market-driven benchmark for the issuance o f private securities. The current capital controls will need to be reviewed to ensure that they do not undulyrestrict regional integration and the movement o f capital across the region. 165. Microfnance industry. Improving access to finance for micro enterprises would also contribute to lessening the constraint imposed by the highcost o f capital in Mozambique. The microfinance industryhas grown rapidly inrecent years, but still has a small outreach, with a highconcentration inMaputo. There are good prospects for its development inMozambique and several best performing micro finance institutions are already inpartnership with well-renowned, international microfinance service providers. However, important bottlenecks need to be eliminated, such as the lack o f human resources.' ' Mozambican micro finance institutions could also strengthen their cooperation with well-known international NGOs. E. IMPROVINGACCESSTO AND QUALITY OFENERGY 166. Electricity i s the most serious infrastructure problem for the Mozambican manufacturing sector, with nearly 64 percent o f firms ranking it as a major or severe problem. Power outages and oscillation o f the voltage has led to the loss o f equipment. As capacity utilization increases, and firms beginto engage incontinuous production, an erratic power supply will become an ever-increasing constraint. The Investment Climate Assessment (World Bank, 2003a) reported that firms suffered 17.5 power outages per month, or on 193 days inthe year, well above the nine African countries surveyed whose average was 77 days.' l2 Outside Maputo it i s more acute: inthe Center and the North, average monthly power outages were 30 and 29, respectively. A study o f cotton ginneries found that power outages reduced the 81-day season by five days and forced ginners to install generators, whence the cost o f electricity i s $O.O85/Kwh, as opposed to $0.035/Kwh on the grid (World Bank, 2005c, p. iv). 167. Residential consumers also face inadequate services. Only 220,000 households (some 6 percent o f all households) had access to electricity. An additional 50,000 connections are targeted for 2005. Since an additional 100,000 households will be established duringthat period, this means that the l o w electrification rate will slightly decrease. This i s explained by the inadequate performance o f the state-owned, vertically `lo arrangementsareprogrammedtostartin2006,permittingrelianceonshort-term(ovemightto7 Repo day) intervention. '''The newDecree 5712004 provides for microfinance companies, including microbanks which can take deposits. There are now no discriminatory restrictions on microfinance companies, provided that they satisfy the minimumcapital requirements laid out inthe Aviso 412005. '*See the cross-country data at htm:!"i.worldbank.orniInvestmentClimate/. 54 CHAPTER4: PRIORITIES FORPRIVATE SECTORDEVELOPMENT integrated electricity company, Electricidade de Moqambique, as well as by the lack o f periodic tariff adjustments for prolonged periods, so that the sector i s unable to generate the necessary revenues for investment. 168. The Government o f Mozambique i s well aware that the lack o f adequate infrastructure, including electricity access and quality, i s a constraint on economic growth. Its short-term response i s to establish the "Beluluane Industrial Park-Free Zone" (BIP-FZ) some 50 kilometers from Maputo, occupying 646 hectares. The works will be completed by end-2005. Useful though it is, this approach i s limited inscope, and a more systematic solution i s needed. 169. Recommendations. Inorder to improve the access and the quality o f energy supplied to residential and industrial users, and giventhat most issues inthe provision o f electricity are related to the lack o f investment and concem distribution and supply, it would be important for the authorities to consider focusing the reform underway on this aspect o f the business duringthe first phase o f the reform. This will help Electricidade de Moqambique improve its efficiency by: (i) intensifying connections where the backbone supplylines are already inplace, and (ii) preparing for a private/public partnership, which should be inplace by the end o f this phase. The latter will enable the Government to leverage additional funding for further intensifying connections and rehabilitating and reinforcing the network, while (iii) taking advantage o f a concessionary aid scheme for otherwise non-commercial rural electrification. This third element will help mitigate the negative impact o f rural electrification on Electricidade de Moqambigue's financial performance and help address the issue o f affordability and accessibility o f services for low-income people as well as for areas where there i s a low population density. 170, Inaddition to these changes at the distribution level, the advantages and disadvantages o f unbundlingo f Electricidade de Moqambique and creating a separate corporate transmission entity to provide transmission assets and perform system operations would need to be assessed, so as to determine whether this approach would contribute to improving the system's overall efficiency. 171. To be effective, these improvements will needto be coupled with enhanced financial, and operational performance at Electricidade de Moqambique. To that end, the level o f energy losses, which i s at more than 20 percent, will need to be brought to a more reasonable level. Since unavoidable technical losses are estimated at 8 to 10percent, there i s still room for improvement. Additionally, the ratio o f payroll to operating expenses also needs to be brought down to intemational standards, while accounts receivable, which have deteriorated to 147 days, need to be improved. 172. Consumer services also needs to improve drastically. Pre-paid services, which were recently introduced, could be expanded as a means o f improving revenue collection and customer services. This would particularly true if 24-hour automatic vending machines were installed to increase customer access to vending points. Electricidade de Moqambique could also launch a comprehensive Guide Book for its customers inorder to provide them with all relevant consumer-relatedinformation. Finally, better management and proper customer notificationprior to power outages will also improve customer satisfaction and help firms plan and manage their production process. 173. These changes at the operational level will need to be coupled with an institutional and legal reform that will aim at establishingthe enabling environment that 55 MOZAMBIQUE COUNTRY ECONOMICMEMOUNDUM i s conducive to the sector's development. Inparticular, it i s important that an adequate regulatory capacity be established, within a transparent environment that will allow the establishment and periodic review and adjustment o f tariffs. This enabling environment will also help protect the rights o f consumers as well as promote renewable energy sources, which could contribute to productivity gains. In so doing, adequate analysis will need to be carried out to determine the advantages and disadvantages o f establishing one or more Multi Sector Regulatory Agency as a mechanism for achieving the desired cost effectiveness and harmonization across sectors. This would include comparisons with alternative mechanisms for achieving the same objectives. F. FACILITATING TRADE 174. Intvoduction and overview. Rapid expansion o f exports has been the main source o f growth duringthe past ten years inMozambique. Since the early 1990s, exports have expanded at 10 percent annually, much faster thanworld exports (6 percent). As a result, Mozambique's exports-to-GDP ratio has doubled. Mozambique has thus become one o f the few countries inAfrica that has seen an increase inits share inworld exports. This rapid export expansion was led by private initiative, both national and foreign. 175. The export-led growth was inducedby moderate improvement inthe trade regime, investment climate, and business environment. The strong response o f exports to these limited improvements inpolicy environment i s a good indication that the private sector i s responsive to changes in the incentive system. Given its excellent natural resource base, its low wages, and favorable location, Mozambique has the potential to maintain its current growth rate. Sustaining this growth rate will require a higher level o f broad-based investments, inaddition to the large-scale projects that generate relatively fewer direct and indirect jobs, compared to labor-intensive manufactured exports. Furthermore, inorder to sustain broad-based growth, there is a need to increase the development of linkages between the mega-projects and local small and medium enterprises. 176. As inmany other countries, for Mozambique, a private sector-led, export-based growth strategy represents the best long-term potential to sustain a high growth rate in light of the existing conditions: the domestic market i s limited, and has a relatively low purchasingpower. Additionally, given that the national saving rate i s not highenough to generate adequate growth to make a marked impact on poverty, a particular emphasis needs to placed on attracting foreign direct investment. Trade performance 177. Mozambique's export performance has been strong, growing at 22 percent annually inUS dollar terms between 1996 and 2003 (Table 25). Export eamings have risen from 6 percent of GDP in 1998 to 20 percent in 2003. Most o f this expansion was due to three mega-projects: Cahora Bassa electricity, the Mozal aluminium smelter, and the SASOLnatural gas pipeline. This expansion i s substantially faster than that o f world See also more detailed macroeconomic variables, including trade variables, in Annex Table 1, p. 135. 56 CHAPTER4: PRIORITIES FORPRIVATE SECTOR DEVELOPMENT exports o f about 6 percent. As aresult, Mozambique i s one of the rare countries inAfrica whose share inworld exports has risen. 178. The "traditional" export base, onthe other hand, is still weak andpoorly diversified, having grown at only 2.3 percent between 1996 and 2003."4 Traditional exports are concentrated ina few agricultural products-prawns, cotton, timber, raw cashew nuts, and sugar. Inparticular, clothing and footwear showed barely no increase: these exports were $5.8 million in 1999, and stood at $6.0 million in2002. The trade liberalization has apparently not helpedthis sector to grow, despite the benefits o f the Africa Growth and Opportunity Act (AGOA) and other trade preferences. Giventhe importance o f labor-intensive export growth for economic development, these discouraging figures put inreliefthe critical need for hrther reforms inthe business environment. Table 25. Exports by main category 1990 1996a 2000 2003 Ann. growth '96-'03 U S $ m i l l i o n s p e r c e n t Exports (f.0.b.) 126 226 360 896 22 Megaprojects 0.0 0.0 127 681 n.a. Traditional exports 126 226 233 363 2.3 Prawn 43 70 69 76 7.1 Cashews (raw and nuts) 14 47 28 29 5.2 Sugar 7.9 13 5.2 19 -7.4 Fishery products 3.9 9.0 8.9 16 8.6 Timber 1.5 9.8 8.8 20 1.7 Miscellaneous 55 78 113 203 -6.2 Sources: Live Data Base, in tum from Mozambican authonties, World Bank staff estimates and IMF. aI995 was not selected becauseexports were unusually low in that year. 179. Imports have grownparipassu with exports. Imports amountedto US$245 million in 1998, and rose to over US$1.2 billionby 2004 (see Annex Table 1, p. 135), or 26 percent o f GDP. A good part of the increase was generated by the mega-projects, particularly during their construction phases.'l5Remaining imports also grew, however, by about 7 percent per year, inpace with GDP growth. The imports o f textile and clothing products in2002 ($20 million'I6) far exceeded exports by the sector ($6 million). 180. Direction oftrade. On the export side, the EUaccounts for about 61 percent o f export earnings (Toyoshima and Yoshino, 2004). The share o f the EUi s rising because o f its absorption o f aluminium. South Africa accounts for 18 percent o f Mozambique's exports (Table 26). Zimbabwe accounted for 20 percent in 1998 but this had fallen to 6 'I4Unrecordedexports, particularlytoMalawi, mayhavegrownfaster, althoughnotime-series data are available. See Macamo (1999). `15 The impact o f MOZAL may be seen in Annex Table 33 (p. 159): alumina imports (the rubric "other metals", which skyrockets in 2001), transportation equipment (which rises in 1999) and machinery. 'I6Seethetable of commodity compositionofimports, inAnnex Table 33 (p.159). 57 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM percent by 2002. Malawi takes $10 million of Mozambique's exports, and the other countries in the region each take less than this (DTIS, p. 1-25). Table 26. Exports by country of destination,1998-2002 (in percent of total exports) I 1998 1999 2000 2001 2002 OECD countries 37.8 37.3 39.6 14.8 51.6 Japan 4.8 4.3 4.3 4.2 0.7 Netherlands 2.3 1.7 1.o 1.o 0.3 Portugal 7.6 9.0 11.6 4.0 4.4 Spain 13.0 12.7 10.7 3.8 2.8 United Kingdom 1.4 1.o 0.9 0.0 0.1 United States 5.7 4.7 4.7 0.98 1.6 Other 3.0 3.9 6.4 0.8 41.8 a Other countries 62.2 62.7 60.4 85.2 48.4 SouthAfrica 17.6 26.2 14.6 15.3 17.7 Zimbabwe 19.5 14.9 17.7 5.3 5.8 Other 25.1 21.7 28.1 64.6a 24.9 Total 100 100 100 100 100 Memorandum item Total exports (US$ millions) 245 284 364 703 682 Sources: Mozambicanauthorities;and IMF, Direction of Trade Statistics. Quoted from Meeuws (2004, Table 7). aBaldednumbers refer to figures that include exports of aluminium to the EU. The statistics do not permit a disaggregation. 181. Onthe import side - see Table 27 - a new element is that the EuropeanUnion (EU)has become the most important partner for Mozambique (over 50 percent), even though this does not emerge very clearly inthe data available (the EUi s included under `Other' in `Other countries'). Correspondingly, South Africa used to be Mozambique's main supplier (57 percent in 1999)' but its role had declined to some 30 percent o f imports by 2002. 58 CHAPTER 4: PRIORITIES FORPRNATE SECTOR DEVELOPMENT Table 27. Imports by country of origin, 1998-2002 (in percent of imports) 1998 1999 2000 2001 2002 OECD countries 31.9 27.2 28.6 20.9 22.2 France 2.1 0.9 2.2 1.1 1.7 Japan 3.9 3.9 4.6 0.6 3.4 Netherlands 0.8 0.7 0.5 0.9 0.5 Portugal 7.9 5.6 7.6 8.4 6.2 United Kingdom 2.0 0.9 1.6 1.1 1.o United States 5.3 6.7 3.5 1.8 4.4 Other 9.9 8.5 8.6 7.9 5.0 Other countries 68.1 72.8 71.4 79.1 77.8 South Africa 39.5 57.2 49.8 40.7 30.3 Zimbabwe 2.3 0.5 0.6 0.8 0.8 Other 26.3 15.1 21.0 37.6a 46.7 Total 100 100 100 100 100 Memorandum item Total imports (US$millions) 817 1,199 1,163 1,063 1,263 Sources: Mozambicanauthorities;and IMF, Direciion of TradeSiatistics. Quoted from Meeuws (2004, Table 9). aAbout half the import data for 2001originate in the European Union. The statistics do notpermit a disaggregation 182. Terms of trade. Mozambique's terms o f trade, overall, declined by 14percent in 1998 and again by 14percent in 1999 (Annex Table 1, p, 135). The aggregate terms o f trade rose by 25 percent in 2001, but the latter i s really an indexing issue arising fiom the introduction o f massive aluminium exports for the first time. Trade policy 183. Mozambique has made important progress in tradepolicy, as was noted in Chapter 2. The trade weightedaverage tariff i s 9 percent (Yagci, 2004), one o f the lowest inAfi-ica. There are four non-zero bands 2.5, 5, 7.5, 25). The top tariff rate has been steadily lowered, from rates approaching 100percent inthe early 199Os, down to 25 percent in2003, and this was accompanied by a rationalization o f rates and the almost total elimination o f quantitative restrictions and export tariffs. Tariffs on capital goods and intermediates are between 5.0 or 7.5 percent, and those on raw materials are 2.5 percent. The differential o f 15-20 percent creates highrates o f effective protection- particularly for processed agricultural products, paint, tableware, and plastic goods. For this reason, as well as to reduce the scope for discretion, the DTIS (2004) recommended that there be a uniform tariff. Furtherreductions intariffs will follow as the SADC trade protocol i s implemented. The plani s to reduce the top tariff rate to 20 percent in2006. Exceptions to the top tariffinclude: a variable surcharge on importedsugar (about 90-100 percent); surcharges on cement and galvanized steel; and surcharges on certain "luxury" items such as cars. It i s notable the many agricultural products, including meat, dairy, and h i t and vegetable products are charged the consumer good duty rate. 184. On the export side, the only export tax that remains i s that on unprocessed cashew (18 percent). As with the sugar surcharge, the authorities do not intendto phase this 59 MOZAMBIQUECOUNTRY ECONOMIC MEMORANDUM export tax out inthe foreseeable future because they feel that protection for cashew processors is needed for the survival o f the industry. The only export ban extant i s a ban on exports o f uncut high-quality woods (see Chapter 5, p. 90, for a discussiono f forestry industry and the export ban). 185. Duty exemptions. Export processing zones may be set up inspecific geographical regions or for specific firms. Eleven o f these had been approved by 2004. The benefits are duty-free imports o f inputs and machinery, tax incentives (a 60 percent reduction in the corporate income tax rate), and exemption from VAT. 186. In2003, the Ministryo f Commerce and Industryattempted to eliminate tariffs on imported inputsrequired for export production, limiting the purviewto large firms (sales o f $250,000) and to certain sectors (textile, clothing and footwear; food processing; other agro-based industries; metal and mechanical; chemicals, plastics and rubber). Only 16 firms had been approved by 2004. Most o f these, however, do not export, and so the effect o f the change has been to increase the effective protection o f domestic producers. l7 187. Tradefacilitation. A recent trade facilitation audit (Meeuws, 2004) identified a two persistent problems: (a) High transport costs. Bothroad andrailway costs are higher than international norms for well-run systems. For example, it costs $7,000 to truck a 24-ton container from Maputo to Pemba, which i s 2% times the cost o f shipping it from Dubai or China ($2,550).'18 (b) Inadequate customsfacilities, lack of harmonization of customs procedures with neighboring countries and lack of coordination among inspections at border crossings. Management o f customs was contracted out duringthe 1990s, leading to increases in efficiency o f collection which more than compensated for the decline inthe tariff rates. Since then, the management o f customs has undergone further reform as the service is being turned over to national operators (Mwangi and D e Wulf, 2003). Waiting times at the borders are still long, however, owing to inadequate border facilities and a lack o f cooperation between customs, other border agencies, transport authorities and law enforcement authorities. 188. Trade agreements. Market access is not a bindingconstraint on export growth at this time, butmaybe a serious constraint inthe medium term. Mozambique has duty-and quota-free access to the European Union (EU) market, under the Cotonou Agreement and Everything-But-Arms (EBA) Initiative; and to the U S market, under the Africa Growth and Opportunity Act (AGOA). Mozambique's access to the South African market i s also relatively free. While the SADC Trade Protocol has very rigid rules o f origin, they are not yet binding for Mozambique. Additionally, Mozambique has access to the Nordic countries (Denmark, Finland, Norway and Sweden), through the Nordic/SADC Accord, See DTIS, vol. 1, p. 30. Compiled by Global Development Solutions, LLC, for World Bank, 2005c, p. 116f. Inturnfrom DTIS (2004) and Sowman(2001). 60 CHAPTER4: PRIORITIES FORPRIVATE SECTORDEVELOPMENT which provides market opportunities for SADC-made products on favorable terms. Mozambique's inability to benefit fully from trade preferences i s due to constraints on the supply side. 189. Recently the question was raised as to whether Mozambique would benefit from membership inthe Southern African Customs Union. The mainbenefit would be duty- andquota-free access to the SACU market without meeting SADC's restrictive rules o f origin. However, benefits o f similar size can be secured through other ways, e.g. using the existing (or an improved) bilateral trade agreement with South Africa, which has more liberal rules o f origin compared to SADC. Furthermore, SACU membership cames two distinct disadvantages. First,byjoining SACU, Mozambique would adopt the present tariff structure o f SACU, which serves largely the interests o f South Africa. For example, South Africa produces some capital and intermediate goods and protects them heavily. Mozambique imports these products and prefers to have zero or very low tariffs on them (Yagci, 2004). Second, trade diversion resulting from SACU membership may be large because SACU's common external tariffs (CET) are highand S A C U accounts for only 26 percent o f Mozambique's exports. On balance, it appears that the benefits have not yet been shown to outweigh the costs, and further studywould be needed to determine whether Mozambique shouldjoin. A reform agenda for economic diversification 190. The broadening o f the base for growth requires the implementation o f second generation reforms to deepen and expand the efforts to establish an enabling environment that i s conducive to broad-based, private sector-led growth. Several recommendations to that end are already made inthis Memorandum. Additional recommendations on the most bindingconstraints that need to be removed are presentedbelow regarding market access, the trade regime, trade facilitation and "behind the border" issues. If adopted, they will contribute to fostering trade development. 191. Inthe mediumterm, market access maybe abindingconstraint, when Mozambique reaches its full export potential. To avoid this, Mozambique needs to actively participate in(a) the on-going, mid-term review o f SADC Trade Protocol to improve the ROO; and (b) the reciprocal Economic Partnership Agreement (EPA) with the EU, which will replace the non-reciprocal Cotonou Partnership Agreement. To this end, technical assistance can be provided to the authorities and the private sector under the IntegratedFramework program, to improve trade capacity inMozambique inboth the public and private sectors and participate inthese negotiations more effectively. For the time being, it appears that S A C U membership i s unlikely to serve Mozambican interests. 192. The trade regime will undergo further change inthe future. As remarked above, the top tariff rate is programmed to fall from 25 percent to 20 percent inJanuary 2006. In addition, Mozambique i s committed to further liberalization through SADC Trade Protocol, and will negotiate a further duty phase-down inthe context o f the EBA. 193. Much needs to be done to improve the support services for exports to eliminate the anti-export bias inherent inthe import duty system. The priorities are: extension o f export incentives to small companies, acceleration o f dutyhax reimbursement, setting up an export credit system, and improving standards and the capacity to meet the sanitary 61 MOZAMBIQUECOUNTRY ECONOMICMEMORANDUM and photo-sanitary requirements particularly for the EUmarket. Technical assistance can be arranged under the IntegratedFramework to meet the priority needs. 194. There i s a need for institutional capacity strengthening for public institutions that play an important role intrade, such as the Export Promotion Agency (IPEX), the National Directorate for Industrial Property andthe National Institute for Norms and Quality or (Instituto Nacional de NormalizaqZo e Qualidade, INNOQ). Conceming the latter, a national policy on quality has been drafted and is awaiting approval by Council o f Ministers. Moreover, a strategic plan for quality improvement has been developed by INNOQ, with the assistance o f Swedish Cooperation. An action plan should be developed to make the INNOQ an autonomous and financially sustainable institution to regulate food, drug and consumer product quality control. 195. Trade facilitation. Inorder to address the delays remarked on above, there i s a need for advance data exchange between controlling government entities and transport operators (e.g. via web sites, border crossing manuals) with full information on tariffs, procedures and necessary documents. There shouldbe further simplification and standardizing o f documentation requirements (viz. no duplication o f intemationalhational documentation). An integratedborder management approach would also involve pre-arrival processing and advance clearance, controls on the basis o f riskmanagement and selectivity, such as to balance security andfacilitation requirements. Further inland clearance facilities should also be built. 196. Behind the border policies: Given that the macro economic and fiscal conditions are improving and the exchange rate regime i s broadly appropriate, the key constraints on the proposed strategy are the cumbersome investment climate and the weak infrastructure system. A selected number o f these questions are addressed inthis chapter. G. CONTRACT ENFORCEMENT 197. The overall quality o f the legal andjudicial system inMozambique i s unsatisfactory. Courts take long to render legal decisions, and the enforcement o f these decisions takes too long. This situation i s partially explained by the large number o f vacancies at the first instance (district) and second instance (provincial) courts. Additionally, there i s a significant number o f sittingjudges inboth the first and second instances who lack the formal legal training to discharge their functions effectively. This situation i s further exacerbated by the fact that a number o f district courts have not been established. As a consequence o f this ineffective and unreliable legal andjudicial system, the cost and risk indoing business inMozambique has increased. Many firms avoid using the legal andjudicial system to resolve their commercial disputes. 198. Public confidence inthe legal andjudicial system i s low. A study on corruption byEtica Moqambique (2001) revealedthat 58 percent o fMozambicans believed that "many" or a "majority" o f court officials were "involved incorruption". While there have been recent efforts to start reducing corruption in a systematic way (Le. with the establishment o f the Anti-comption Unit and enactment o f the 2004 Anti-Corruption Law), the public confidence has continued to erode due to a number o f highprofile scandals. 62 CHAPTER4: PRIORITIESFORPRIVATESECTORDEVELOPMENT 199. The 2005 Doing Business Report finds that it takes, on average, 540 days to resolve a commercial dispute inMozambique. This compares unfavorably with South Africa (277), Lesotho (285) and Botswana (154). An important factor inthe highcost o f external finance and lack of trade credit inMozambique i s the difficulty o f enforcing contracts. Efforts should be directedtoward improvingthe efficiency o f the judicial system and encouraging the widespread use o f the private system for Alternative Dispute Resolution establishedin2000 to help enforce contracts. 200, Recommendations. An effective legal andjudicial system requires sufficient institutional capacity and resources to properly manage the system and make it fully functional to renderjustice and have it enforced in a timely manner. 201. The Ministry o f Justice, the Supreme Court, the Administrative Tribunal, and the Office o f the Attorney General couldjointly carry out an assessmento f the skills o f their current staffto determine whether such staff would be able to effectively carry out their functions and provide training, where needed, to develop staff skills. Additionally, it would be important for the respective institutions to consider implementing an intemal Performance assessment system for their staff. This would allow the management to assess the timeliness and quality of the performance o f individual staff and their respective contributions to an appropriate performance by the legal andjudicial system. Among other features, the performance assessment system could include an incentive system which rewards staff who carry out their functions professionally and honestly, and penalizes those who do otherwise. 202. This approach could prove effective ifit also includes the design and implementation o f analytical methods and data collection o f court cases to monitor the performance o f the legal andjudicial system, including areas that may require specific attention, Inorder to bring about these desiredchanges, the respective institutions mentioned above would also benefit from developing and implementing a hiring and training program to attract and retain quality key staff to work as court administrators and be appointed in other critical positions inthe legal andjudicial system. 203. Given that the formal judicial system i s overburdened, the authorities could consider strengtheningalternative disputeresolution mechanisms. The Arbitration, Conciliation, and MediationCenter (CACM) is an independent, non-profit organization which was established to assist inresolving commercial disputes. Since becoming operational in2002, CACM has taken up and resolved fewer than 10 cases. The authorities should consider innovative approaches to expand the dispute resolution services inorder to expedite legal decisions and accelerate their enforcement. H. "QUICK WINS" Reducingthe time and cost for registering a business 204. Current status. Overall, the Mozambicaneconomy has been slow to adjust and reallocate resources to their most efficient use. The difficulty o f entering and exiting the market has reduced competitive pressure to improve productivity and allowed inefficient firms to survive. It has also discouraged informal f i r m s fiom becoming formal and availing themselves of some o f the benefits of formality. It i s thus important that the 63 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Mozambican authorities move swiftly to bring the time and cost o f starting a new operation inline with its international competitors as a way o f easing entry of firms inthe formal economy. To that end, the registration procedure must be simplified, expedited, and made more transparent. Moreover, all public organizations that play a part inthe registration process should provide better support to investors when going through the firmregistrationprocess. 205. Two surveys report that it takes between 132 and 153 days to start a company in Mozambique (2003 Investment ClimateAssessment Survey and the World Bank publication, Doing Business in 2004: Understanding Regulation). While any data on the numberofdays and cost that it takes to register a company inherently has a degree o f subjectivity and could thus be disputed, there i s a general consensus inthe business community that the registrationprocess o f firms i s extremely complex, time-consuming and costly. The regional average i s 63 days. InMozambique, it costs 96 percent o f the per capita incometo register a firm. Few entrepreneurs arewilling to endure thisprocess and pay such a highcost, so entrepreneursmay give up on the investment or remain unregistered. 206. The registrationprocess i s cumbersome andtime-consuming. For example, the step o f commercial registration, which takes a few hours inmany other countries, took over 30 days. Additionally, the publication o f the company bylaws takes well over 30 days. 207. Recommendations. The Government could consider two sets o f complementary actions. The first one i s the streamlining o f the registrationprocess and integration o f the operations o f all three institutions involvedinthis process. I t i s recommended that the three institutions, namely the Notary Office, the Public Commercial Registry (Consewatdria de Registo Comercial) and the State Printing Office (Imprensa Nacional) merge into ajoint Management InformationSystem so that the relevant public officials o f any one o f these institutions can retrieve the information entered by another institution, process it and make it available for processingby yet another institution involved. 208. The second set o f actions i s to modernize the registrationprocess by computerizing the entries and retrieval o f information regarding the registry on enterprises. Currently, the steps to register a company are the same as those usedinthe 19'h Century, when the Notary Public wrote out the articles of incorporation in longhand inledgers andprepared handwritten extracts for Public Commercial Registry andthe Printing Office. Modernizing this process would allow the authorities not only to quicken the speed at which enterprises are registered, but also increase the accuracy o f the registry. Inaddition to this, proper consideration o f the outsourcing o f the publication function also needs to be assessed; and, if appropriate, the publication should be sourced out to private parties. Approvingthe new CommercialCode 209. Current status. Mozambique's Commercial Code dates back to 1888. (Johannes Brahms was composing his masterpiece Op. 117.) This Code i s inneed o f a substantial revision so as to modernize the law governing commercial transactions. Relatively new commercial concepts, such as goodwill or leasing, are not taken into account inthe Commercial Code. Commercial disputes could be avoided by having a clear Commercial 64 CHAPTER 4: PRIORITIES FOR PRIVATE SECTORDEVELOPMENT Code, thereby relieving the overburdenedjudicial system, which takes 540 days to resolve the average business dispute. Information technology could be usedwith legal backing. 210. To remedy this situation, the authorities have beenworking on a revision of the Commercial Code since 1998. Although a revisedCommercial Code was submitted to the Parliament in2001, it has yet to be approved. It seems likely that, with a close follow-up on the work of the technical team, the Commercial Code, revised to the satisfaction o f the Government, the private sector, and the Parliament, could be completed inthe last quarter o f 2005. 211. Recommendations. Regardingthe process, it is important that the authorities ensure that all the relevant stakeholders are adequately consulted and that a consensus i s established regarding the amendments to the Commercial Code. Giventhat a Commercial Code has been submitted to Parliament and that it i s generally accepted to be adequate, theprocess leadingto the approval of anew Commercial Code bythe Parliament could be done through the establishment o f a tripartite technical team, which would include representatives o f the Government, Parliament andthe private sector. Accelerating value added tax refunds 212. The Mozambican VAT refundpolicy "Cddigo do Impost0 sobre o Valor Acrescentado (Cddigo IVA)" was established to provide tax relief and stimulate activities for eligible private sector firms and activities which fall under the value-added tax refund structure. However, delays inVAT refundsto private firms have become a significant constraint to private sector development inMozambique. Inthe sample o f 194 firms surveyedfor the preparationo fthe 2003Mozambique Investment ClimateAssessment,24 firms reported beingowed refunds, and the average amount owed was more than 13 percent o f their annual sales. This situation contributes to further straining the cash flow of these firms ina country where the cost o f capital i s highand considered a binding constraint on private sector development. One firm was awaiting a refund o f over 3 billion Meticais (approximately $150,000) for over 12 months. 213. While some delays intax refundscan sometimes be attributedto an inadequate understandingo f VAT refundprocedures by private operators, this factor does not explain the delays experienced by firms that have gone through the process several times and have a sound accounting system. Several firms interviewed suspect that the delays in VAT refunds are due to cash shortages inthe VAT refundsystem. A potential solution i s to permit offsetting the V A T refundagainst upcoming VAT payments for imports (offsetting i s already allowed inrespect o f sales made within Mozambique.) 214. Another dimension o f this issue i s that some firms are not compliant with their VAT payment. The authorities should more effectively pursue firms that are non- compliant in order to increase revenues to pay compliant firms. The private sector associations should also encourage firms to be compliant with their tax payments by ensuringthat theywork with suppliersthat are also VAT compliant. 215. Recommendations. The time it takes to make VAT reimbursements should be lowered to 30 days. To that end, the authorities will need to adopt several measures: P Hiringadditional VAT staff at the DirectorateofTaxes and Auditscouldalso be considered to improve collections. 65 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM P All the relevant laws and regulationsshouldbedisseminated. P A task force couldbe established to review all cases and ensure that the respective refund payment i s processed and made. Reducingthe cost of inspectionsand fines while improvingcompliance 216. Inorder to ensure compliancewith relevant laws and regulations, the Mozambican authorities undertake inspections related to public health, environmental safeguards and protection, fire safety, labor standards and safety. Firms registeredand operating inMozambique can thus be the object o f these inspections. While most companies understandthe need for these inspections, there are concerns about the way the inspections are carried out and the resulting cost for their respective enterprises. The main concerns that entrepreneurs raise are as follows. 217. Generally, the inspections are done ina random, inconsistent and uncoordinated fashion by many authorities and for different purposes. Some firms are frequently inspected, while others are not. The unannouncednature o f inspections is disruptive. A labor inspection can take place at a time when the responsible staff for humanresources i s away from the office, and the inspector requires the concerned staff to stop working and provide any set o f documents, some dating back a few years. 218. What concems entrepreneurs most are two other characteristics. The first one i s that it seems that the inspections are not aimed at bringing about compliance with the existing laws and regulations, but rather finingcompanies. The second one i s the discretionary manner inwhich most inspections are conducted. Given the absence o f guidelines andor standards on why, how and when to conduct an inspection, the system is characterized by arbitrary and inconsistent decisions. 219. Several entrepreneurs have also complained that the inspections are often carried out by a corrupt, low-level official who assess a high fine on a firm so that the company would have enough incentive to negotiate with the inspector for an "off the record" settlement. These complaints about the existence o f corruption i s corroboratedby the USAID-funded study, Assessment of Corruption and Red Tape as Barriers to Trade and Investment in Mozambique (December 2002), which establishes the existence o f a relatively highlevel o f corruption and the estimated cost associated with it. Based on this study, Mozambique forgoes between US$64 andUS$137 Million o f FDIdue to redtape and Corruption. Additionally, Mozambique also loses revenues through "off the record" negotiatedsettlements o f fines between inspectors and some entrepreneurs. More importantly, small informal businesses do not have the incentives for entering the formal sector and small- and medium-size enterprises stated that they were the most negatively affectedby the inspections. Finally, this type o f system has the perverse effect o f allowing dishonest entrepreneurs to evade compliance with existing laws and regulations, while corrupt inspectors receivebribes. 220. Two additional areas o f weakness further compound the ineffectiveness o f the inspectionsystem. The first one i s that inspectors seldom provide either a rationale for the fine they have assessed or how the amount o f the fine was determined. The second one i s that the current inspection system does not distinguish between firms which have a good track record on being compliant with existing laws and regulations and those that have a poor record of compliance. This discourages firms which comply and maintain a 66 CHAPTER4: PRIORITIESFORPRNATE SECTOR DEVELOPMENT good record from being compliant and reflects a lack o f focus on firms where there i s real need for follow-up inspections. 221. The Ministryo f Industryand Commerce (MIC) has started to emphasize the need for inspections to aim at educating the entrepreneurs and bringthem into compliance with the existing laws and regulations. While this initiative will certainly contribute to slightly reducing the cost and inconvenience o f the inspections for businesses, it will only be effective if and when the same initiative i s broadly adopted by other ministries, such as Ministryof Planning and Finance, Ministryo fHealth, Ministryo f Labor, Ministryo f Tourism, and Ministryo f the Environment. 222. Recommendations. The system o f inspections needs to be simpler and better coordinated across the several ministries that undertake inspections. The system would be accepted and effective only if it i s transparent, consistent, fair, and allows for appeals. 223. From an implementation standpoint, it would seem wise to avoid wholesale changes inexisting administrative requirements. Instead, efforts should be aimed at bringing about changes inpriority areas and incrementally. Some o fthe "quick wins" that could be pursued are as follows: P Inspectionsshouldbe announced with at 15 days' noticeto allow the firm to prepare documentation and plan its activities to minimize disruption to the work flow. Upon completion o f the inspection, the firm should be given an official, written document stating that an inspectionwas carried out. P Eachministryshoulddevelop rules and standardoperatingprocedures for carrying out inspectionsto significantly reduce discretion and ensure consistency. 9 To minimizeharassment anddiscretionarybehavior, inspectors shouldpreparea written report containing their findings. P Fines should only be levied following the issuance o fa letter to the firm stating the nature o f the violation as well as the corrective measure(s) to be adopted to become compliant. P All fines shouldbe made payable to the Director o fTreasury at the Ministry o f Planning and Finance. This would ensure that collection and payment procedures for fines levied be registeredcentrally and completely, separate from individual ministries. 224. The ministries may need to carry out an assessment o f the skills o f its current staff to determine whether such staff would be able to effectively carry out their functions and provide training where needed to develop staff skills. Additionally, it would be important for the respective ministries to consider implementing a internal performance assessment system for inspectors which will allow the respective ministries to ensure adequacy o f fees, provide written reports, and strengthen the system so that preferential treatment i s not given. To this particular end, the assessment system could include an incentive system which rewards inspectors who carry out their functions professionally and honestly, and penalizes those who do otherwise. 225. The private sector associations also need to help inimproving the inspection arrangements. This assistance could take the form o f seminars to disseminate rules, 67 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM regulations andprocedures related to the inspections to their respective members, once such rules are established. Table 28. Summary of businessclimate constraints and proposed recommendations Type of ad- Main characteristics Some proposed solutions Proposed ministrative implementation barriers period Limited 2% revenuelosses due to Intensify power connections Starteda few years access and power outages Improve operator's performance ago quality o f 18 average monthly power Will continue on for electricity outages the mediumterm Cost of and Low level of credit (18% of Strengthenbanking supervision Started2001 access to GDP) Enhance financial inter- Will continue for the finance Highinterestrate (>= 30%) mediation medium term Highlevel o f collateral Improve monetary andpublic (141% of credit amount) debt management Slow, costly Enforcementis slow: 540 Broadenthe use of alternative Startedmid-1990s contract days dispute resolution mechanisms Will continue on for enforcement Costly and risky for firms Create specializedcourts for the medium term commercialdisputes Highcost and 138 days to register a firm Streamlinethe registration Startedin2004, time to re- Cost: 96% ofper capital process 2005: main gister a firm income Modernize the registration implementation process 2006: consolidation Commercial Current Code is 116 years old Adopt the Commercial Code Startedin 1999 Code Absence of modem concepts Use a consultativeiparticipatory 2005: periodof Code inadequacies Inadequatefor modem trade processto do so. approval 2006: consolidation period Slow value 99 days averagerefund time Use offset mechanismto refund 2005: periodof main addedtax Severe constraint on firms' Disseminate,clarify existing actions refunds cash flow laws 2006: consolidation Hiremore inspectors period Establisha task force for refund High cost of Random, inconsistent, Scheduleinspections 2005: period o f main inspections frequent Emphasizeobtaining com- actions Too punitive, not educative pliance Will continue for the Rewardgood firms, penalize medium term firms that are non-compliant 68 CHAPTER 5. THE ROLE OF NATURALRESOURCES IN FASTER GROWTHAND POVERTY REDUCTION 226. Natural resources are central to this Memorandum on growth and poverty for three reasons: 9 Mozambiqueis a naturalresourcebasedeconomy. Agriculture is thebiggest contributor to growth and to poverty reduction, and other resources are also crucial - fisheries (10-15 percent o f exports), tourism and mining (mega-projects). This chapter seeks to show that effective naturalresource management will be crucial for maintaining the process o f growth inthe future. There are important growth opportunities inforestry and mining. Land reforms could present growth opportunities inthe form o fmedium-scale and large-scale capitalized farming. 9 The poor are crucially dependent onnaturalresources andchanges inlegislation can dramatically affect their welfare. It i s crucial to understandthe mechanisms o f the present allocation systems so that when policy changes are introduced, the welfare o f the poor i s notjeopardized. Also, without further reform some o f the MillenniumDevelopment Goals will not be attained (e.g. inwater). 9 There is aparticularity about the legal setup: generallyitis focused so determinedly on equity ofaccess that less attention was given to the specific mechanisms for ensuring access, to promotingan efficient use o f resources, and to monitoring and enforcing compliance with the norms. The challenge becomes one o f how to secure growth opportunities without reducing equity o f access. 227. Hence this chapter starts by examining the characteristics which the five key natural resources have in common, specifically inrelation to their role in growth creation and poverty reduction. Then i t proceeds to examine each o f the five key natural resources intum, explaining their potential contributions to sustaining highgrowth rates andreducing poverty further. A. NATURAL RESOURCES: CHARACTERISTICS INCOMMON 228. The five key natural resources o f Mozambique share certain characteristics. One i s that their legal and administrative frameworks were, without exception, designed with the intention o f protecting the poor, but that various factors conspire to result inoutcomes that frequently fail to protect the poor. A second i s that the Government is failing to collect rents, at a considerable fiscal loss. The third i s that there i s frequently a gap between the written regulation and actual implementation - the gap arising from low capacity, excessive interference, and rent seeking behavior - and this creates uncertainty for the poor, as well as for investors. The fourth characteristic i s that there i s frequently a lack o f transparency inprocedures; this reduces accountability and encourages arbitrary and rent-seeking behavior. 229. Equity in design, but ofen inequitable or sub-optimal outcomes. Natural resources management policies inMozambique suffer from a fundamental general MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM problem. The laws rely on the use o f instruments for the management o f natural resources which have proven successful inmost OECD countries as well as inmany developing countries. This includes long land leases, water basin committees, delegated management of drinking water supply, quotas, concessions and fees infisheries, forestry and mining. Inthe cases o f all five key natural resources, the poor have benefited from intentionally free or nearly free-access regimes, which have been serving as economic buffers. Nonetheless, while much attention has been given to ensuring equitable access to resources, less attention has beengiven to the speczjc mechanisms for ensuring access, for promoting an efficient use o f resources, and also for monitoring and enforcing compliance with the norms. Between the Constitution, the laws, regulations, and final implementation and enforcement there i s a myriad o f factors which frequently lead to poor outcomes. 230. The framework laws and regulations seek to protect the poor: inthe case o f land, fisheries, forests andmining, regulations focus primarily on larger scale, commercially oriented private interests, while poorer populations benefit from simplified procedures to access resources, or simply have complete open access. The problem i s therefore not the access to these resources but the incentives to make optimum and efficient use o f them. The economic implications o f their potential gaps and failures are very significant, particularly inthe case of land. 231. Inmany cases, the regulationsappear to have been"captured" by government bureaucracies which have created unnecessarily complex procedures that may deviate, in implementation, from the original intentions o f the framework laws. Opportunities for rent seeking and corruption have been created, harming bothprivate commercial interests and the poor. 232. Ineconomic terms, while the framework laws are clearly designedto protect the poor, and this i s a laudable achievement since poverty i s the key issue inMozambique, open access to natural resources cannot be the solution (except inthe case o f subsistence fisheries, para. 308, p. 89). The next challenge i s to implement an effective regime o f secure and transparent property rights that respects the interests of traditional users but permits and encourages more intensive development o f resources. Such development will require the combination o fnatural resources with substantial capital investment, much o f which will have to take the form o f foreign investment. Hence, property rights mustbe seen as being secure for both local populations andforeign investors. 233. Failure to collect rents. Major efficiency and equity gains could be achieved in the management o f four o f the resources by implementing better mechanisms for charging for both services and access to resources. At present, the government i s foregoing significant rents for access to and the use o f resources without achieving any clear benefits inretum. 234. Subject to the inevitable uncertainty about such estimates, public revenues from natural resource rents could be increased from US$30 million to US$67 million per year under current economic conditions, mostly by adopting a reasonable level o f rentdcharges for land (Table 29). Looking ahead over the next decade, ifthe policies recommendedinthis study were adopted, public revenues from natural resource rents in 2015 could realistically increase to U S $ 215 millionper year, representing almost 3 percent o f GDP or 20 percent o f total tax revenues. The bulk o f the increase after 2005 70 CHAPTER5: NATURAL RESOURCES will come from the miningsector, where investments are already expandingvery significantly, so that the total contribution from this sector alone would almost equal that of the other natural resources combined. Table 29. Currentand potentialrents from naturalresourceuse Current rents P o t e n t i a l r e n t s US$ m i l l i o n p e r y e a r , c o n s t a n t p r i c e s 2003 2005 2015 Fisheries 24 37 40 Forests , 3 3 5 Mining 3 7 120 Agricultural land 1 20 50 TOTAL 30 67 215 As a percentage of GDP' 0.7 1.5 2.9 ASa percentage oftax revenue' 5.3 11.0 19.4 ' Basedon CEMprojections See the relevantsectionin this chapter for the mechanismsresultingin the increased rent capture 235. Gaps between regulations and their implementation, generating uncertainty. The way regulations are written andimplemented createsmajor opportunities for rent seekingby the government bureaucracy. There i s excessive government interference in part originating from the fact that all resources belong to the State, introducing an unnecessary level o f uncertainty, bureaucracy, and an unfavorable business environment. Inthe areaofmonitoringlicenses, leases, etc., where the government should have a strong grip, it shows weak capacity. This lack o f capacity to monitor inturn creates opportunities for interference and rent seeking. The current regulations and their actual implementation have been leading to both inefficient use and capture by elites, contrary to the objectives o f the overall framework laws. 236. Many examples support the above conclusions, as i s shown inthe remainder o f this chapter. The transfer o f land titles requires the land to have been developed in accordance with the authorized landuse plan. The government capacity to actually monitor this i s minimal, and inenforcing the regulation there i s great scope for bureaucratic discretion and thus either corruption or meddling, creating an uncertain business environment."' Inthe case offisheries, there is a complex combination of quotas, license fees and payments for royalties which do not apply uniformly to all the industry,but are differentiated inaccordance with criteria which are obscure. It is difficult to trace revenues, licenses, allowed catches, and other critical economic parameters o f the sector. This hinders the effective collection o f government rents and monitoring the extent to which the industryis incompliance with regulations. 237. Efficiency and equity gains would be reinforced by clarifying many regulations that create uncertainty, introducing a higher degree o f transparency inthe system, and 'I' Teams arebeingsetupinthe provincesto inspectthe landuseplans. 71 MOZAMBIQUE COUNTRYECONOMIC MEMORANDUM improving the capacity to monitor licenses. Also, it i s essential to minimize the numerous opportunities for rent seeking and corruption which undermine more efficient and equitable use o f natural resources. 238. Lack of transparency in procedures. Clear and transparent rules are very basic requirements o f a good business environment. They are also critical to make bureaucratic and political bodies socially accountable and thus to minimize rent seeking and corruption. Natural resources inMozambique are characterized by low degrees o f transparency. For instance, inthe concession o f licenses and equivalent permits inall sectors, the poor communities are effectively alienated from the process o f resource allocation. Transparency i s lacking also inthe publication and dissemination o f various forms o f data and information: catch statistics, land rights and land transactions, fines imposed and collected in all sectors, forestry management plans, and others. 239. For instance, inthe case o f lands, rights to large areas o f land are beingacquired or transferred on the basis o f expectations about its value infuture uses, while the land itself i s not being farmed or used for other purposes. Inthe case o f fisheries, licenses and concessions are granted under procedures that lack transparency, a problem that i s acknowledged even by the fisheries administration. 240. The subsequent sections on the five key natural resources will discuss how these problems - poor outcomes, low rents, uncertainty and lack o f transparency - may be addressed. B. RURALLAKD 241. It was found inChapter 1that the agricultural sector was the second-largest contributor to growth inthe past decade; this coincided with the findinginChapter 3 that agriculture was the sector that contributed most to reducing poverty. But growth inthe agriculture sector has slowed and technological advance has been limited. Therefore considerable efforts will be needed to ensure continued poverty reduction. Among the efforts the Government can do i s to improve the administration o f two o f the key natural resources: water and rural land. Inthe section on rural water (p. 75) it will be shown that selected reforms could help bring about increased irrigation access and reduce risk to agricultural households from floods and droughts. Inthis section, o n rural land, it will be shown that land access for smallholders i s reasonably secure but that for large-scale endeavors is not. While the Government's effort inincreasing agricultural productivity prospects (p. 1lo), it will be shown here that at negligibly low cost the large-scale should concentrate primarily o n smallholders, as i s argued inChapter 6 on growth commercial sector could be further encouraged to invest also, and this can help reduce poverty by creating employment inrural areas. 242. The total land area o f Mozambique, excluding rivers and inland waters, i s about 784,000 km2.The FA0 estimates that about 360,000 km2(36 million ha) are cultivable, but the area cultivated for arable and permanent crops was estimatedto be only 4.9 million ha in2003. Over 90 percent o f the 4.9 million hectares o f cultivated landin2003 were devoted to food crops. The total area cultivated for food crops increased by only 0.9 percent per year from 1997-98 to 2003-04. The pressure on cultivable land i s relatively low. 72 CHAPTER5: NATURAL RESOURCES 243. Projections suggest that the growth inlanduse for crop production and mixed farming area can easily be accommodated within the existing land resources over the next 10-15 years. Cropping rates on cultivable landwill remainbelow 20 percent for the country, while total land usedfor cropping andmixed farming systems should remain less than 40 percent up to 2020. This should meanthat there i s ample scope for providing land for a rapid expansion o f medium and large scale commercial farming that has not beenincludedwithinthese estimates. 244. Land distribution. The estimated total number o f farm holdings i s inthe range 3.1 to 3.2 million inthe period 2000-03 with an average cultivated area of about 1.35 ha per holding and a total cultivated area o f about 4.25 millionha. Holdings with more than 5 ha o f cultivated land account for a little under 12 percent o f the total area under cultivation. Since farms o f 50 ha account for less than 2 percent o f cultivated land, this group does not represent a threat to small-scale farming. 245. Legal and institutional arrangements. The 1997 Land Law, together with subsequent regulations, sought to resolve land conflicts by including several core principles: P All landremains property ofthe state, but land leasescanbe granted for upto 50 years. They are renewable, inheritable, and transferable subject to prior administrative authorization. 9 The awardofaleaseis subject to thepresentationofalanduse or development plan, which can be cancelled ifthe lessee fails to comply with its conditions. 9 Investmentsininfrastructure andimprovements onleasedlandcanbebought and sold, again requiringan administrative authorization. 9 Traditional landuserights are recognizedandformalized through a systemof community landmanagement, with co-titling o f community lands. P Local consultation is requiredbefore land leasescanbeawarded. The law succeededinits main aim ofprotecting the traditional land use rights o f smallholders. Inmany respects the new land regime established under the law has been a success: agricultural production has beenincreasing, some investments are being attracted, there are no landless peasants and no tenants are payingrents to absentee landlords, as i s observed insome other African countries. 246. However, Mozambique has still not succeeded inattracting large amounts o f capitalized and commercially oriented farming investment, and that few partnerships betweenlarge landusers and local communities have taken place. Capital-intensive operations continue to complain about the difficulties o f obtaining secure leases and navigating the bureaucratic system. This suggests that further reforms are needed which go beyond the provisions o f the 1997 Land Law and its accompanying regulations. (Obviously there are also other constraints on investment, such as inadequate infrastructure, as elucidated inChapter 4 on private sector development). 247. Related to this problem i s the result that on account of the way inwhich land i s allocated, many o f the concessions are under-used. Since the system o f acquiring land rights lacks transparency and allows the well-connected to obtain large holdings at virtually no cost, rent-seeking behavior i s encouraged. The existing mechanisms for the 73 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM award o f land rights cannot be expected to deliver either efficient or fair outcomes in current circumstances. Indeed, it could be argued that ifthe prospective profits from land conversion are large, a bureaucratic system cannot control the process o f land acquisition or transfer other than by completely suppressing all manifestations o f market transactions. A combination o f greater transparency and more attention to market incentives is required to produce outcomes that are seen as being reasonable and capture a share o f the surplus generated by development for the State as landowner. 248. This is not an argument for a freehold land arrangement. It i s proposedhere to retain the leasehold arrangement but to make it secure and untrammeled by bureaucratic interference. To summarize the reforms which will be discussed inthe following paragraphs: 9 The leasehold arrangement is leftintact. Nochange to the law or the constitution i s envisaged. Rural leaseholds o f land will still not be saleable as such. However, improvements and buildingswill continue to be freely transactable. P The landtax onproperties over 10hais raisedto, say, $12 per hectare, which is a little over half the level which a free rental market would have thrown up. 9 Bureaucraticinterference isreducedby abolishing landuseplans andbyletting rural land automatically accompany the sale o f improvements andbuildings, as i s done with urban land presently (at least intheory). It is argued that this set o fmodest reforms will alter the system o f incentives such as to attract investment by small, medium and large-scale farmers, while protecting smallholders' rights. 249. The rationale for the land tax is as follows.'2o The land law regulations specify a standard tax rate o f Mt 30,000 per hectare but with a series o f multipliers according to holding size, location and use, so that they range between Mt 7,500 and Mt 180,000 per ha. Customary land users are exempted. The elaborate differentiation o f tax rates i s too complex for consistent implementation. Above all, the tax rate i s extremely low and i s the tax is not consistentlypaid.'21 The total revenue from land taxes o fMt 13 billionis equivalent to only 0.07 percent o f agricultural value added. Hence it has no impact on land use or land distribution. 250. Ifthe Governmentwere to establish theprinciplethat landtitles will be subject to a land tax, then future and current land occupants would build this expectation into their decisions regardinghow much land to occupy and how much to invest inimprovements. Ifthe enhancedlandtax were imposedonlyonholdings over 10haofcultivated land, and assuming that the share of imputed rent to value added i s 10 percent (a low figure by comparisonwith other countries), then the imputed rent would be Mt 55-60 billion or US$18-20perhectare of cultivable landper year. This level ofimputedrent could readily permit a land tax of, say, $10-12 per hectare. Data do not exist to permit complex differentiation of the tax by land quality. Since the objective is inany caseprecisely to 120 Note that the proposal for an effective land tax is not new. The recommendation was made in 1993 during the Bank's agricultural missions, and was emphasized in the Agricultural Sector Memorandum (World Bank, 1997). 12' As o f 1999-2000, about 50 to 60 percent o f concession holders paid the tax. 74 CHAPTER5: NATURAL RESOURCES encourage high-value uses rather than low-value uses, the best way to do it is to pitch the rate on the l o w side, compared with putative rents, as has been done inthe above computation, and impose it per unit area without any differentiation. The likely revenue of a little over US$lO million i s not large, but it would establish a clear principle that land taxes must be paid on land leased fiom the government inlieu o f rent. Holders o f large leases would, intime, either put their land to better use or cede the unused parts o f their leases. 251. The next recommendation i s to eliminate bureaucratic interference, or the threat o f it, as it i s a disincentive to invest. Scope for bureaucratic discretion, and thus either corruption or meddling, occurs when leases are transferred. The mainpressure point concerns whether or not the land has been developed inaccordance with the authorized land use plan. It would be much betterto rely upon an appropriate set o f economic incentives for the efficient use o f land resources than on bureaucratic monitoring. Penalties and enforcement for non-compliance with conditions on land leases must be simple, unambiguous and easily enforceable. Inpractice, this means that payment o f the land tax i s the only condition that can reasonably be enforced. 252, The requirement that potential lessees propose and implement a development plan for the land is theoretically understandable, but the government does not have the resources to check the validity o f and subsequent com liance with development plans for commercial agricultural land throughout the country.lf2 Hence it i s proposed that this requirement be dropped for rural land in all but very special exceptional cases. This would imply that rural land leases would be able to accompany sales o f improvements and buildings automatically, as leases are (theoretically) permitted to do inthe case o f urban land.'23 This would allow transfers to be entirely a private matter subject to registration at the National Land Registry. 253. Havinginthis way eliminated the scope for bureaucratic interference when people seek to transfer land titles, the leasehold title system should provide all the security that is needed by investors at the medium-scale and large-scale level. C. WATER 254. Water i s essential for poverty reduction and improvement o f quality o f life in Mozambique. As will be shown inthis section, ifthe country is determined to improve its poverty indicators and to reach the Millennium Development Goals, it should place a great deal o f emphasis on the sound management and use o f its water resources, and seek to focus more on the provision o f improved water supplies to the rural population and to poor groups inurban areas. Additionally, Mozambique needs to expand its provisiono f water for smallholder subsistence irrigation -which has an important role to play in poverty reduction - and commercial agriculture, and reduce its vulnerability to droughts. 122 Note that families and communities are not required to present land use plans. 123 Note that there i s no legal or administrative impediment to the sale o f improvements and buildings. But, as pointedout above, inrural areas it does not automatically follow that the landlease onwhich the improvements and buildings is located is also transferred to the new owner. For the land lease itselfto be transferred, a further administrative authorizationmust be sought. It i s here proposed that this administrative requirementbe scrapped. 75 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM 255. Water availability. Mean annual rainfall i s around 800 - 1,000 mm along the coast, around 1,200 mm inthe mid-part o f the country, and between 1,000 - 2,000 mmin the North. There i s considerable variation - even within the wetter Norththere are areas o f low rainfall. Upstreamrivers are important for water resources: 46 percent o f surface water resources are generatedby local rainfallwhile the remainder originates inupstream countries. Some 14million Mozambicans -nearly 75 percent o f the total population- rely on groundwater supply. Wells have an average depth of 50 m, allowing for the use o f handpumps. 256. Imtitutional and Zegdframework. The institutional and legal framework for the water sector inMozambique i s overall coherent and is largely consistent with experience and good practice inmany middleand high income countries. The management structure was established with the approval o f the Water Law in 1991. The Ministry o f Public Works and Housing(MOPH) has the central role inwater management, with cross- sectoral coordination ensured by the National Water Council (CNA), a consulting body o f the Government. Within MOPH, the National Directorate o f Water (DNA) i s in charge o f water planning, and the A R A s (Regional Water Administrations) for water management. Following the publication o f the National Water Policy in 1995, responsibility for implementation resides with a number o f (relatively) autonomous regional and sectoral entities. 257. Inthe areaofwater resourcesmanagement, de-concentration anddecentralization are beingimplementedwith the creation o f Regional Water Administrations (ARAs). In the area o f water supply and sanitation, the 1995 Water Policywas approved and started to implementa framework for delegated management, attracting the participation o f the private sector for utility management. Under this framework, two new institutions were created: the Water Regulatory Council (CRA) which i s responsiblefor economic and other regulation of water sector systems that are under delegated management, and the Investment and Assets Fund for Water Supply (FIPAG), which has been set up as the body that owns the assets inurban areas that are either managed or leased by private operators. For all other urban areas as well as for the rural water supply, DNA continues to assume the leading role, until autonomous institutions can manage all systems in the mediumterm. 258. Public expenditures in water. Total public expenditures inthe water sector were betweenUS$17 and 19 millionbetween2002 and 2004, equivalent to between 0.3 and 0.5 percent o f GDP (Table 30). The water supply and sanitation sub-sector have historically accounted for between 83 and 97 percent o f total expenditures inthe water sector, the remainder being allocatedto water resource management. Government expenditures in sectors other than water were much higher, education and health each getting over 4 percent o f GDP, roads getting around 3 percent and agriculture 1percent. 76 CHAPTER NATURAL 5: RESOURCES Table 30. Expendituresin selectedsectors, 2002-2004 Expendituresas percentage ExpendituresinUS$ o f GDP I 2002 2003 2004 2002 2003 2004 Total expenditures 30 29 27 1,082 1,249 1,479 of which: Education 4.4 5.1 4.9 158 220 270 Health 4.1 4.1 4.1 146 176 230 Agriculture 1.3 1.o 1.o 46 42 55 Roads 2.8 3.8 3.5 102 165 195 Water 0.5 0.4 0.3 17 18 19 Memo items Total expenditures(Mt trillions) 25.6 29.7 33.4 GDP (Mttrillions) 85.2 102.8 125.3 CPI (1998=100) 148 168 189 Exchange rate (Mt/US$)i1000 23.7 23.8 22.6 Source:Budget ExecutionReports. 259. Insightinto the relative distribution o f spending onwater supply as between urban and rural areas comes from the joint Bank-Government Public Expenditure Review o f September 2003 (World Bank, 2003b),'24 as well as from updated spending data for 2004.'25 Overall spending on urban water has historically exceeded that o n rural water. N o w we factor inthe fact that the ruralpopulation i s 73 percent o f the total, and the fact that rural people are much less served (27 percent) with water from improved sources than are urban people (65 percent). It emerges that Government spending (own revenues and donors' contributions) on expanding water access, per head o f those as yet unserved, was six times greater for urbanthan rural in 2003 and 2004. It should be noted, however, that projects to the value of $135 million are being developed by the African Development Bank and bilaterals includingDfID and SIDA. It may be that the tide has tumed infavor o f concessional/grant financing o f rural water supply. 260. Water resuurces infrastructure. InMozambique, there are only five large dams with sufficient storage capacity to have an impact on multi-year uses including flood control. The useful capacity o f the country's dams represents only 5 percent o f the mean annual runoffof the country's rivers, excluding the Zambezi. This i s very l o w by 124 The PER usedthe Government's PARPA (2001) and the Revised PARPA Water Program Budget preparedby the National Directorate of Water. See FinneyandKleemeier, 2003 (Table 9), for sources and detailed discussion. This backgroundpaper was preparedfor the 2003 Public ExpenditureReview (World Bank, 2003b). 12' These data originate with the Departamentode CooperaGZo Intemacional(DCI) of the Ministry o f Finance. The data sets assembledby DCIare used (together with accounting data) to compile the quarterly Budget ExecutionReports. 77 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM intemational standards and indicates that the country needs to develop its infrastructure to store a larger share o f the runoff.126 261, The Government could make significant progress inreducing vulnerability to droughts by dealing with issues relatedto rural water supply, irrigationwater for smallholder farming, and the operation and maintenance o f existing systems. This could be complementedbymeasures improvingthe information database, includingwith upstream riparians, and waming systems that would serve both for the management o f droughts as well as floods. Small dams and reservoirs should be built throughout the country inthe most drought-prone areas. 262. Floods and droughts have direct implications for poverty. Just to quote one example: with the cyclone o f early 2000, some 11percent o f the cultivated landinfour provinces was flooded, and about 490,000 people, most o f them rural dwellers,were either displaced or trapped in flood-isolated areas (World Bank, 2002, p. 4). The important subject o fnatural disasters and poverty i s not dealt with here because it has beencoveredthoroughly ina contemporaneous study, The role ofwater in the Mozambique economy: identifiing vulnerability and constraints to growth (World Bank, 2005b). The study sought to quantifythe cost of water shocks, both at the macroeconomic level and at the household level. It notes the coping strategies employed byhouseholds - game hunting,sale o f firewood, temporary labor migration, and so forth. It recommends reducing the vulnerability o f the ruralpoor to water shocks through flood management and increasingwater accessibility for productive use. 263. Storage reservoirs are certainly the most controversial o f all structural measures, especially inlight o f potential environmental and social impacts. Dams can contribute to the attenuation o f floods andthe mitigationo fdroughts but they can also aggravate the effects of floods ifnot well maintained. Two major dams have beenidentifiedas being urgently needed-Moamba Major, for the water supplyo f Maputo, and BueMaria, for the water supply o f Beira and for the large irrigation developments being considered for the Lower Pungoe. Medium sized dams are also requiredinthe short to mediumterm to ensure the water supply to cities such as Nampula, Nacala, Quelimane and Lichinga. This will be a major effort for the Government and it will require strong support from the donor community. 264. The operating rules for all existing dams should be reviewedand adjusted to current needs interms o f flood attenuationas well as other uses. As most floods originate inupstream countries, Mozambique will have to continue to work with upstream riparian countries to ensure that the operation o f upstreamreservoirs does not contribute to aggravating floods in its territory. Implementingthe legal, institutional and regulatory framework will involve capacity development, institutional strengthening, and the development of water resources management tools such as basinplans, mapping and contingency plans. 265. Irrigation. About one-tentho f the country's 36 millionhectares o f cultivable land i s suitable for irrigation. The total area o f land currently under cultivation i s about 4.3 million ha. Of this only 0.9 percent i s irrigated, the lowest inthe region (compare 126 Excluding Cahora Bassa, the storage per person i s 330 m', far lower than South Africa at 746 m', See furtherdiscussionof the storage backlog inWorld Bank (2005b). 78 CHAPTER 5: NATURAL RESOURCES Kenya at 1.6percent and Zimbabwe at 4 percent). The benefit-cost ratio for small-scale irrigation i s high: an investment o f US$20 million per year over 15 years would increase total agricultural value added by about US$ 250 millionin2020 (at 2005 prices), raising the overall rate of growthof agricultural value added from 6 percent to 6.7 percent per year. 266. Ina context oflimitedwater storagecapacity, andalack of financial resources, public resources would better be concentrated on improving conditions for smallholder farming than on commercial irrigation. For cost effectiveness, the focus would best be on small-scale, low-cost schemes rather than large projects. Attention should be concentrated on the best located and least expensive sites, rather than large capital- intensive schemes. Encouraging smallholder irrigation would also leadto increased high- value crops and increased incomes for some of the nation's poorest households. Smallholders will needto form associations inorder to be able to afford irrigationand win access to technology and markets, thereby moving from subsistence to small commercial farming. 267. While there i s substantial potential for the expansion o f commercial irrigation, this should be drivenby the private sector. The role o f the Government i s to put inplace the legal, institutional and regulatory framework, as well as to support the development of water storage infrastructure provided this i s linked to multi-purpose projects (energy/ water / transport / drought management). For reasons o f controlling risks and mobilizing the necessary investment resources, the private sector i s likely to be much more interested insmall or medium-sizedschemes ratherthanthe largerprojects that tendto attract international donor interest. Hence, irrigation inthe Zambezi, Punge, Buzi, Limpopo and Incomati basins i s likely to be most attractive for commercial farmers, since the irrigation systems would rely primarily upon divertingwater from rivers and small dams. 268. Some support for the development o fwater storage infrastructuremay be warranted when this i s linkedto multi-purpose projects through energy production, water supply, transport, or drought management. Private-public partnerships should also be considered, for example with the Government providing funds for storage works and rehabilitation o f irrigationinfrastructure and the private sector reserving part o f the irrigated area (10 percent for example) for small farmers and giving them support for production and commercialization. Irrigated sugar-cane production, for example, has potential to foster development by means o f outgrower arrangements. 269. Ifcommercial and smallholder irrigation are to contribute to longterm agricultural growth, it will be necessary to promote investment not only inimgation infrastructure but also incomplementary activities ranging from transport to export processingand marketing.'27 Irrigation i s only one element in a larger package that i s currentlyhinderedby all o f the issues o fthe business environment. Without the market environment and infrastructurethat meets the needs o f commercial agriculture, there will 127 Managing the risks o f market development i s also a critical issue. Investment in sugarcane irrigation, estates and mills to expand commercial production i s a heavy commitment, while the world market i s notoriously volatile. The Government should not promote such investments unless it i s able to attract private partners willing to share the risks. Similar considerations apply to the development o f commercial and smallholder schemes focusing on crops such as citrus, tobacco, tomatoes, other vegetables and fruit that have substantial processing or marketing requirements. 79 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM not be the skills, processing plants, etc required for smallholders to make the transition to higher value-added activities. 270. Water supply. Of all the natural resources, the one needing the most attention bypolicy-makers is water. The lack o f cleanwater is a major factor inpoor health and lowered work productivity. Eventhough great strides have been made inmaking and implementing far reachingpolicy reforms, given the huge challenges that the sector represents, a tremendous amount remains to be done, particularly inrural water supply. 271. Here we build on work done by the Public Expendituve Review (World Bank, 2003) - informed by the dramatic finding, which was not available for the latter study, that rural water access i s lower than had beenthought (namely 27 percent as opposed to 40) and urbanwater access is higher than had been thought (64 percent as opposed to 38). The detailed data are as follows. 272. According to the household surveys (IAFs), access to clean water inrural areas was raised from 12 percent in 1996/7 to 27 ercent in 200213, and inurbanareas from 56 percent in 1996/7 to 64 percent in2002/3.'2' The Millennium Development Goal'29i s to halve the proportion o f people without safe water by 2015, which means raising access inrural areas to 56 percent'30and inurban areas to 78 percent. Clearly there is a much greater distance to be traversed inrural water (27 to 56) than inurban (64 to 78). The PARPA objective for urban water was 50 percent coverage of householdpiped water by 2005; this goal was obviously unrealistic and has not been attained (the level i s approximately 30 percent). Urgent improvement inwater access i s called for. This will also require major investments. 273. Urban water supply. The quality o f urban water service does not compare well with countries at a similar level o f development. Interms ofunaccountedfor water, Mozambique has a median o f 48 percent inthe 5 biggest cities (Maputo: 53 percent) in 200413', exceeding the African average o f 39 percent.'32 Businesspeople complain that 12' The Government's official figures o f water access are 40 percent for rural areas and 38 percent for urban areas, These sets o f numbers are not necessarily incontradiction; they measure different things. The IAF reports the outcome, namely whether the householdhas access to an improved water source as per the MDG definition (see the definition at www.who.int/water sanitation healthlmonitoringiinip042.pdf). The Government's number i s a computation o f the number o f boreholes dug, multiplied by a coefficient to allow for breakdowns, and multiplied by a coefficient for the number o f households usingthe average borehole, This Memorandum uses both sets o f numbers, but will focus the MDGdiscussion on the household survey number- because: (a) the same basis is usedby the household survey inboth urban and rural areas, (b) household survey numbers are internationally comparable, which the administrative records (output-cum-coefficients) are not because the coefficients vary from one country to another, and (c) it i s an outcome measure rather than an input or output measure. 129 For a summary o f all the M D G s inMozambique, see Annex Table 2, p. 136. The base year for all the MDGs is 1990. W e assume that the figures for 1990 and 199617 were the same. 131 Data from the Government (June 19, 2005). See also: Aide-mCmoire o f the Second National Water Development Project, World Bank, August 23-September 23,2004. 132 African average UFW: see World Health Organization and UNICEF, Global water supply and sanitation assessment 2000 Report, section 4.4. Source: http:!!wuw.who.int'docstoreiwater sanitation health;GlobassessmentiCilobal4-4.htm. See also Kirkpatrick et al. (2004, p. 42) who report the African average for 1999-2001, for state-owned water 80 CHAPTER5: NATURAL RESOURCES water supply failures occur 152 days in the year, the worst o f nine African countries surveyed, which together avera ed 56 days. 133 Hours of service vary from 11to 19 hours inthe five biggest cities,Ik while the average for developing countries i s 20 hours per day, and the average for African cities was 17 hours per day in 1999-2001.135 Staffing ratios are 16 to 24 per 1,000 connections, vs. 17/1,000 for East and Central Africa and 19/1,000 for South Asia.136Bill collection ratios are between 70 and 88 percent inthe five biggest cities.I3' The low levels o f operational efficiency reflect poor management and incentives, skill shortages, and the usual range o f problems that afflict most publicly-operated utilities inAfrica. The PARPA objective for urban areas was 50 percent coverage o fhousehold pipedwater by 2005; this goal was obviously unrealistic andhas not beenattained (the level is approximately 30 percent). 274. The costs o f expansion o f supply differ sharply by method. The network cost o f a new household connection is about US$ 250, while upgradingwater treatment and transmission capacity would double this figure to $500; dividingby the official number o f 5 persons servedper connection, the per capita cost i s $100. The cost o f a standpipe is about $5,000, which covers 25-50 households, or about $27 per capita. 275. Recent new directions inurbanwater supply have turnedout to be a success story.I3* As o f 2000-2001, efficiency was low as was pointed out above. The government opted for a "delegated management'' model, introducedby competitive biddingfor five major cities including Maputo. A private sector contractor commenced operations in 1999. The improvement inoutcomes by 2004 have shown that the right decision hadbeenmade. InMaputo city, and supply improved from 9 to 13 hours, the number o f functioning standpipes doubled, the collection ratio rose from 61 to 73 percent, andthe number o f connections rose from 70,000 to 90,000 (Table 31). Furthermore, the urbanwater sector has attracted further investment. Ifthe arrangement works, the Government intends to contract out the management o f the urban water utilities for five more citiedtowns inthe medium term. utilities, as being 34.8 percent. The data are outdated, but probably the African average has risen since, so that it i s safe to say that Mozambique's cities perform poorly by comparison. 133 Supply failures: see cross-country data from the investment climate surveys at on investment climate at htt~:!irru.worldbank.orp/InveshnentClimate/ . 134 Maputo 13 hours, Beira 14, Quelimane 11, Nampula 19, Pemba 13, as o f M a y 2005. 135 Kirkpatrick and Parker (2004, p. 42) report 17 hours per day o f piped water in African cities for the year 1999-2001. 136 Developed countries have 2.1 staffper 1,000 connections; the best 25 percent o f developing country utilities have 5 or fewer staffper 1,000 connections (Tynan and Kingdom, 2002). The average staffing ratio inAfrica in 1999-2001 for state-owned water utilities was 20.1 per 1,000 (Kirkpatrick and Parker, 2004, p. 41). 13' See also: Aide-memoire o f the Second National Water Development Project, World Bank, August 23- September 23, 2004, which indicates a median o f 65 percent. 13' See a more detailed description o f the institutional change in urban water, inChapter 6, section (c) Futuregrowth and institutions, p. 107, and also paragraph 370, page 109. 81 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Table 31. Water indicators for Maputo city, 2000 and 2004 2000 2004 Connections 70,000 90,000 Functioning standpipes 264 438 Average hours o f supplylday 9 13a Collection ratio (%) 61 73 Source:Walker (2005). * As of May 2005; see above paragraphs for source. 276. Rough estimates suggest that 90 percent access to piped water in all 21 urban centers inMozambique could be attained by 2015 at a cost o f some US$20 million a year. An investment program o f this scale could be financed, provided that the utilities charge tariffs that are designed to cover the full costs o f service. A part o f the cost o f the program could be raised from initial connection charges. When determining the level o f the charge, the authorities should take into account the trade-off between erecting a barrier to achieving highcoverage levels inpoor areas versus the need for adequate fundingfor further expansion. To cover capital costs o fUS$ 500 per connection, it would be necessary to charge an average tariff o f about US$0.55/m3. Infact the average price actually paid per cubic meter by domestic consumers at mid-2005, inthe five major cities, was between US$ 0.42 and 0.51/m3, so these tariffs still need to be raised further. The tariffs are much lower inthe smaller cities (US$ 0.32 to 0.38), calling for increases o f as much as 50 percent. Inaddition there i s a connection fee o f about $100. 277. Ideally, almost all urban consumers would receive service through private connections, but a period o f at least two decades will be required to get close to that goal inMozambique. Hence, standpipes will play an important part inurbanwater supply provision in the poorer areas o f Mozambique's towns and cities for a decade or more. Since the poor are especially dependent on standpipes, the improvement o f standpipe service would make an important contribution to urbanpoverty alleviation and should be given highpriority, as i s already recognizedby FIPAG. 278. The record o f managing urban standpipes inMozambique i s poor. It i s reported that no more than one-fifth function properly. This i s a common pattem in many cities in Asia and Africa where standpipes are operated by municipal or public utilities, which typically have no incentive or resources to ensure that standpipes are maintained and operated properly. Further options should be explored further: (a) standpipes are transferred to community organizations (church groups, local NGOs) which take on the responsibility o f operating and maintaining them, or (b) standpipes are franchised to private operators, who compete with local water vendors. There have been limited experiments inMozambique -notably inTete and Angoche -with both models and these could be extended more widely. Another option (c) i s that where the management o f water utilities i s transferred to the private sector, the incumbent firm be givenproper incentives to maintain and extend the network o f standpipes. Appropriate provisions are being included inmanagement contracts and similar arrangements for private participation inwater services in South and South-East Asia. This option could be considered when the existing arrangements inMozambique are reviewed or extended. 82 CHAPTER5: NATURAL RESOURCES 279. Given the rate of urbanpopulation growth and the existing level o f urban water coverage, it i s unlikelythat 90 percent coverage would be reached within the next two decades ifpolicy were to focus exclusively onprovidinghouseholdconnections. Even if standpipes are an interim and not entirely satisfactory solution, nonetheless they provide a route to ensuring that almost all o f the urbanpopulation has access to some form o f pipedwater supplywithin a few years. 280. Rural wafer supply. Inparallel, the government has pursuedthe "demand driven approach" inrural areas, which requires communities to make a specific request for the installation o f awater source, pay a fee, and take responsibility for its maintenan~e.'~' Even so, at any one time about 28 percent o fwater sources are non-functional for lack o f spare parts close to communities, poor installation or supervision o f works, weak community management and extreme poverty. 28 1. Rural waterpoint installation i s expensive comparedwith neighboring countries. Sinking a borehole (with ahandpump) costs $10,000 inMozambique, versus $5,000- 7,000 inMalawi, Tanzania, Zambia and Ghana. Giventhe country's low population density, the average number o fpeople servedby each connection is about 25O,l4O making the per capita cost some US$50. This highfigure i s mainly due to drilling costs. Some of the differential has been attributed to reasons beyondthe control o f the authorities, such as aquifer conditions, an as yet undeveloped drilling industry, the small size o f the Mozambique market which limits the gains from economies of scale, poor roads, large distances, inadequate infrastructure and services, a highlevel o f business risk associated with highinterest rates, and delayed payments on Government contracts. However, several countries inthe region also face these problems. Additional factors may be poor organization and a lack o f competition in drilling reinforcedby reliance uponparastatal drilling companies. The elements needto be investigated and ifnecessary acted ~ p o n . ' ~ ' 282. Evenassuming no improvements inefficiency, the investmentrequired to meet the MillenniumDevelopment Goal o fhalvingthe proportion o fthe ruralpopulation without access to clean water would be about US$ 130millionover 10 years. Financing an investment program o f US$15 millionper year should not be a major barrier. Costs would be lower ifthe cost o f borehole drilling couldbe reduced.'42 It i s essential that any increased program o f capital expenditure should be accompanied by mechanisms designedto ensure the long-term sustainability o f ruralwater points, including increased support to community-managed schemes and/or the testing o f alternative management approaches such as handpump leasing or private sector management contracts. 13' See Ministry o f Public Works and Housing (2001), p. 21. I 4 O 250 people per well i s the figure used in most international calculations. The Government's computations typically use 500 (see Finney, 2003). Assuming a rural population of 12 million, with clean water coverage o f 27 percent, and 8,000 functional wells, and assuming that half o f rural people get their water from other sources, then the number o f people per well i s 203. 14' A first step towards addressing this issue has been taken with an ongoing consultancy service to Assess National Private Sector Drilling Capacity For Rural Water Supply inMozambique. This study will also evaluate the structural cost o f boreholes inthe SADC region and provide an insight into the major problems facing the drilling sub-sector inMozambique, and find ways to make the sub-sector more competitive. 142 Other forms o f saving can also be made, e.g. hand dug or augured wells, and rope pumps. 83 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM 283. Thepolicy andinstitutionalframework has a weak link: the lack o f clear policies for rural water management, which should encompass drought mitigation, irrigation development and rural water supply. There are targets for selected items among these, such as the rehabilitation o f non-operational irrigation schemes or the expansion o f rural drinkingwater supplies, but there is little sense o f a coherent strategy towards a set o f issues that are critical for a large part o f the population, including most o f the p00r.l~~ 284. One reason for this problem i s the division o f responsibilities between ministries (primarily the Ministry o f Agriculture and the Ministry o f Public Works and Housing). Another factor i s that donor efforts, especially with respect to investments inrural drinkingwater schemes, have beenpoorly coordinated and have, therefore, failed to promote a common approach that could achieve lower costs and better sustainability o f projects. A final aspect, overlaying the other two, i s that this i s a sphere that requires major investment, either from Government or external sources. However, the lack of domestic resources combined with the absence o f a coherent strategy has meant the funds have been used opportunistically rather than with the aim o f achieving specific medium or longer term goals. 285. The underlying difficulties will not be resolved easily or quickly. But, equally, the country is unlikelyto meet the MDGs for water supply or poverty alleviation without higher spending on rural water supply, drought mitigation and smallholder irrigation - either viajoint schemes or specific projects. A major barrier to such investments is the highlevel o f costs exacerbated bythe poor recordo foperationalmaintenance. Puttingin place reliable sources o f fundingcombined with institutional changes to promote competition and greater efficiency inthe provision o f rural water infrastructure will be critical elements o f any policy strategy. 286. Smallpipedsystems. The performance o f small piped water systems i s poor, with most o f them ceasing to operate after a few years. The costs o f rehabilitating or constructing these schemes are much higher than borehole supplies - some as much as $85 per capita. These differentials suggest that the Government should consider the rehabilitation and expansion of small piped systems only when unit costs o f investments are not too high(say, not more than US$50 per capita), when technical and financial sustainability i s ensured for operation and management, management is taken by an autonomous entity operating on commercial principles, and there i s a clear commitment o f the local authorities. 287. Sufficient funds need to be available to cover operation and maintenance expenses. Fewer than 40 percent o f small town piped systems were operational in2001, It is wasteful to install pipedwater systems that are not properly maintained, but equally there i s an opportunity to make a significant step o f improving coverage by rehabilitating defunct systems. A new management model i s soon to be approved by Government which call for private sector participation in the management o f small piped systems. This strategy is expected to provide more sustainability to these systems andto help develop the services ina more professionalmanner. Presentlythe National Water Policy (NWP) of 1995 is under review involving all stakeholders.The revisedNWP is expected to provide for reinforcement o f institutional coordination, long-term sustainability and highlights water as a cross-cutting issue and its role inthe poverty alleviation, economic development including the water resources for irrigation, drought mitigation, water supply and other sectors. 84 CHAPTER5:NATURAL RESOURCES 288. Swifier expansion of rural water supply. Despite the evident progress made, persistingwith the present arrangement for rural water i s inadequate. Whereas inhealth and education Mozambique frequently equals or exceeds the Africa average, inrural clean water access it i s well behind, at 27 percent, compared with Africa's 46.5 percent (see Table 2, p. 4). Gross primary school enrolment i s now over 100percent, and vaccination coverage 82 percent, so the "missing link"inrural livelihoods i s now the absence o f clean water. 289. An integratedstrategy i s needed inorder to meet the PARPA objectives and the MDGs, and help people escape the serious health consequences o f unprotected water. It should be a policy o f swvt and massive expansion of access to clean water. It should maintain two principles: cost-effectiveness, and the "demand-led approach" - for absent the latter, the proportion o f non-functional wells will quickly soar beyond its present 35 percent level. It i s essential that any program o f increased capital expenditure be accompaniedby mechanisms designed to ensure that proper provision is made to ensure that spare parts are available as required, and that community capacity i s created. This process will require institutional reforms, which implies changes inpolicy, capacity buildingat all levels, the development o f a soundprivate sector (consultants, drilling contractors, spare parts dealers, etc.) 290, This mammoth expansion i s within reach. It can be attained by P Increasing resources and personnelinrural water access. Although the approach will continue to be demand-driven, the bindingconstraint presently i s not communities that satisfy the criteria but govemment staff, capacity building and funding to permit it to be rolledout. It may be worth considering shifting resources to rural water from sectors of lesser priority. P Investigate thefactors underlying high drilling costs. Inparticular, consider whether subjecting more work to tender andhaving less done by parastatals would reduce costs. 9 Expandstandpipe accessinurbanareas.144Thiscanbedone at lower cost thanby expanding householdtap connections. One possible management model i s subcontracting to individuals (as i s being done in some Mozambican towns such as Angoche) or a pilot management model inMaputo. 9 Licensingprivately operatedwater kiosks as is done inother African cities. > Increase co-payment in urban areas. Inall cities, and particularly the smaller ones, the tariffs need to be raised so as to cover operation and maintenance and make a contribution to debt service. 144 See Chapter 4, "Standpipes: an evolvingapproachto public water supply", in Plummer (2003) which gives examples insub-SaharanAfrica o f managementarrangements for standpipe water access. 85 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM D. FISHERIES 291. Fisheries are Mozambique's second largest single export after aluminium, accounting for 10-15 percent o f all exports.145Fisheries constitutes about 1.5 percent o f GDP. The largest fishery business inMozambique, by far, is the SofalaBank shrimp fishery (industrial and ~emi-industrial).'~~Artisanal and subsistence fisheries are important for subsistence, with about 90,000 persons active. Fishproducts account for over 20 percent o f the animal protein supply. These two broad categories - shrimp and artisanal-cum-subsistence - are the focus o f the discussion here, for brevity's sake. 292. Inanutshell,the keyfindings are two: 0 Management o f the commercial shrimp fisheries i s inneed o f improvement for the sake o f fairness and increasingthe tax take, but not so much for economic growth because the sustainable catch has already peaked. 0 The artisanal and subsistence fisheries do not need any reorganization or policy changes at the present time. They continue to be crucial as a poverty buffer. 293. Shrimps. The landed shrimp catch i s some eight thousand tons annually, for a value o f US$70 million. The total rent from the shrimp fishery i s o f the order o f US$24 million. The sector i s dominated by three companies ("industrial"): Pescamar, Efripel and Krustamoz, which together holdtogether more than half o f the total shrimp quota. In addition there are some 200 smaller companies with licenses; this "semi-industrial" sector consists ofjoint ventures -joint with Emopesca, a parastatal. The value o f the company has been estimated to be inthe range o f US$lOO million, but i s potentially much larger. Inaddition, a land-based processing industryhas emerged, as it was declared compulsory for the industrial fisheries to land their catches regularly and not use long-term refrigeration. The processingcompanies are vertically integratedwithjoint venture companies. 294. Fishingeffort has increased by a factor o fthree since the late 1970s, although the total catch has remained nearly constant. The total catch, broken down by industrial and semi-industrial groups, i s depicted inTable 32. 145 . Fisheries usedto account for between 35 and 40 percent o f exports inthe 199Os, prior to the expansion o f exports arising from the mega-projects. 146 Smaller ones (together accounting for just 7 percent o f the value o f the total catch) are tuna inthe Mozambique Channel, kapenta in Cahora Bassa, and gamba (deep water shrimp). For brevity, only the Sofala Bank shrimp fishery and artisanal-cum-subsistence fishing are discussed inthis Memorandum - the former being most relevant for economic growth and the latter as a poverty buffer. Further detail may be found inthe backgroundpaper by Margulis (2005). 86 CHAPTER5: NATURALRESOURCES Table 32. Quotas and catches -shallow water shrimp trawlers with freezers, 1995-2003 I I Quota2)(t) Catch (t) industrial Semi- Semi- Year Industrial Industrial') industrial') Total') Total') Total3) 1995 7,344 157 7,501 8,615 1996 7,043 396 7,439 8,123 8,183 1997 7,462 8,239 514 8,753 9,605 9,825 1998 7,650 7,172 976 8,148 8,559 8,559 1999 7,940 1,645 6,97 1 1,474 8,445 8,475 8,806 2000 7,750 1,505 7,419 1,72 1 9,140 9,420 9,429 2001 7,735 1,140 7,730 1,566 9,296 9,479 9,401 2002 8,000 1,165 9,222 9,472 2003 8,260 1,025 7,990 'Eide et al 12003) MdP, 2004 'Total catch ofPenaeus shrimps in Mozambique, FAO, FishStat2004 295. Untilvery recently EfripelandPescamarwere subject to a specific taxation regime relative to their exploration o f natural resources (Taxa de ExpZoraqGo dos Recursos Naturais, TERN). The companies did not pay license fees but paid more than US$3.5 millioninTERN in 1999. Recent information indicates that the two companies will now pay the normal license fees, which amountedto around US$2 millionin 1999, excluding the TERN, which i s now eliminated. 296. Though it i s difficult to estimate the resource rent collection, the net income generatedby the fishery i s estimated to be at least 40 percent o f the total revenue. Assumingthat 80 percent ofthe export catchis produced efficiently, the rent would reach an annual value o f aroundUS$24 millionbefore tax (a corporate tax o f 35 percent o f the accounted profit). This implies that each ton i s generating a net revenue o f around US$3,800 (US$1,900 to the localpartner) plus the license and other fishing fees and taxes paid to the Ministry of Planning and Finance (particularly on fuel). Inaddition, at least one directjob i s created per 15 tonnes o f shrimp caught. 297. No public information i s available on the amounts collected from thejoint ventures by Emopesca and the tax amounts passed on to the Ministry o f Planning and Finance. The share o f the net revenuekept inMozambique by some o f the quota-owners i s likely to be higher than that providedby the joint ventures through Emopesca. Unfortunately, this revenueis not always duly collectedby the Ministryo f Planning and Finance. 298. Managementpolicies, implementation and governance. The Ministry o f Fisheriespromoted in2003 a revision o f the fisheries legislation with emphasis on the Marine FisheriesGeneral Regulation, which was adjusted to be more in line with the current realities o f the sector. Interms o f organizational structure, the separationbetween the political and management components, which implies the establishment o f an independent control entity and the clarification o f its coordinating responsibilities i s the most urgent problem to be addressed. 87 MOZAMBIQUE COUNTRYECONOMIC MEMORANDUM 299. Both industrial and semi-industrial shrimp fisheries are regulated by Total Allowable Catch together with application o f Individual Catch Quota, both based on research conducted by the Fisheries Research Institute. Regulatory measures on the demersal species and large pelagics taken by the semi-industrial fleet are practically absent, except for issuance o f fishinglicenses. Quota management plays an essential role inthe mainMozambican fisheries, particularly inthe shrimp and the tuna fisheries. Since there i s closed season regulation inplace, there i s no danger o f resource depletion. In both cases the quotas, which are non-transferable, function more as a control o f input (fishingeffort) than control o f output (catch), so they could be replacedby ordinary license fees. 300. The priority given to the semi-industrial sector inthe Master Plan combined with some lack o f clarity on its definition have reduced the overall profitability o f the shrimp fisheries. Quotas were assigned to new entrants not by reducing the industrial fleet participation but by increasing total participation inthe fishery. The allocation procedure o f new licenses was not transparent, with substantial values being transferredwithout clear regulation or distribution criteria. Inan attempt to minimize the ensuing political problems, the Ministry o f Fisheries was created in 1998. 301. Even though new semi-industrial trawlers were allowed to install fkeezing facilities, violating the initial objectives o f developing a land based processing industry, increased capacity for on shore processinghas been developed over time. Quality control problems related to on shore production have been addressed and significant improvements have been obtained inthis area recently. The investments instrengthening the semi-industrial sector have nonetheless allowed a landbasedprocessing industryto emerge and to professionalizethe market operations inboth the domestic and regional markets, even though their share o f the total shrimp exports is still insignificant. 302. However, the quotas were assigned to new entrants not by reducing the industrial fleet participation but by increasing the total participation inthe industry, whence the vast increase in total effort. Part o f the reason for the promotiono f the "semi-industrial" fleet was to encourage the entrance o f Mozambicdnational operators. Inany case the arrangement has not been as successful as had been planned at the start, inthat the Mozambican operators have not been able to develop their share and national operators remain scarce. 303. Cost recovery and rent capture. It was a part o f the Fisheries Sector Master Plan to achieve cost recovery for the fisheries administration by 1996. Estimates indicate that full cost recovery has been achievedifthe TERNwere included inthe calculation. A fund allocation systemthat covers the costs o fthe fisheries administration is inplace, currently providing a surplus (40 percent o f the license revenues) to the Ministryo f Finance. Ten percent o f the licenses share i s allocated directly to fisheries institutions. The Fundo de Foment0 Pesqueiro (FFP), which receives and redistributes 50 percent o f the license income, finances the rest o f the operating costs o f the fisheries institutions. Surprisingly, the Ministry o f Fisheries claims not to have any information on the TERN (collected by the Ministry o f Finance), although it has now apparently become a regular license fee as indicated earlier. 304. Interms o frent capture, Mozambique is inthe unique situation o fhaving fish resources which allow for super normal profit (resource rent) to be collected, although 88 CHAPTER5:NATURAL RESOURCES only a small fraction of this rent goes back to the Mozambican government. The license fee as o f 2000 was a very low $220 per quota ton, or roughly 2.5 percent of sale value.'47 The distribution of benefits by Emopesca, a governmental holding company, has not been possible to trace. Its net assets and shareholder funds represent nearly US$ 6 million. 305. Recommendations are proposed intwo areas: regulation andpoverty alleviation, Concerning regulation, licenses and concessions are granted by procedures that lack transparency. The problem i s acknowledged by the Ministry, and the lack o f clear criteria for quota distribution i s beingreexaminedwith a view to implementingmore transparent procedures, as this will simplifyroutines and responsibilities. One barrier seems to be lack o f capacity and staff training. Another problemi s the fragile balance obtained with the many interests involved, which makes any change difficult and only possible with a highdegree o fpolitical authority, which does not seem to be developed at this stage. 306. Resourcerents are obviously beinggarnered by major operators and are leaking out o f Mozambique, eventhough market solutions such as licenses or quota auctions couldbe usedto retrieve at least some o f it. It would, however, be necessary to thoroughly analyze possible negative consequences causedby asymmetric information and financial conditions. Thejoint venture companies are largely controlledby foreign companies, which are the actual operators. A key issue is to develop a national Mozambican capacity to exploit fishresources profitably, and inso doing to review the use of Emopesca as a governmental control instrument. After the introduction o fjoint venture companies inthe shrimp fishery, Emopesca was the majority share holder, while today it i s a mere 30 percent, with little control by the Ministry of Fisheries. 307. I t i s proposed that inthe future a modified form o f the "two-track" arrangement be used. International access couldbe arranged best by an auction system. This would considerably increase the total rent receivedby the Mozambican state, o f which only a small share i s beingretained inMozambique at present. The reason for not advocating the "first-best" efficient method o f an auction system throughout i s that this would likely exclude national operators almost altogether. Hence for national access, a clear set o f publicly available criteria for quota distribution is required, to ensure predictable and fair systems for users, and also to secure social benefits from natural resource exploitation. 308. Poverty alleviation and the role of artisanal and subsistencefisheries, There are an estimated 90,000 persons active inartisanal and subsistence fisheries on Mozambique's coasts.'48 The vast majority o f these people are poor. Fishresources have played an essential role as a poverty reducing economic buffer. General community development has also been initiated as parts o f fisheries development, with assistance from the Instituto de DesenvolvimentoPesqueiva de Pequena Escala. Artisanal fisheries are subject to light regulation, and it i s not proposedto change this. Subsistence fisheries enjoy open access, which i s the best approach with this fishery; it would be important not to underminethe poverty buffer through improper management or other governmental 14' The quota fee was Mt 3,600,000iton. The exchange rate was Mt15,860 = $1. Value o f shrimp per ton inEuropeanmarkets was $9,000. Data from the then Vice-Minister o fFisheries, S.Exa.Alfred0 Massinga. 148 Instituto de Desenvolvimento Pesqueira de Pequena Escala, August 2, 2000. 89 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM interventions. There i s no indicationthat open access inthis case has been leading to depletion o f the resource. E. FORESTRY 309. Forestry accounts for 2.5 percent of GDP presently. The output of the sub-sector i s about 120,000 cubic meters per year. Experts put the potential at some 500,000 cubic meters. The proper exploitation o f this naturalresource i s essential for two reasons: first, it has significant growthpotential'49and can hence provide an avenue for further fiscal contributions, as i s shown below; and second, it provides a poverty buffer.15@ 310. When compared with 1997/98 figures, the current total output level has decreased by almost 15 percent (Table 33). Most o fthis reductionwas due to the recent export ban o f first-class round wood, since it completely dominates the totals. However, processed wood also declined duringthe late nineties and only by 2003 had it managed to reachthe levels experienced in the mid-nineties. National total wood output i s only around 127,000 m3/yeartoday, round wood accounting for 93,000 m3,sawn wood for 30,000 m3, and posts, plywood and veneer for the rest. Table 33. Nationalwood production (m3)1998-2003 Wood Types 1998 1999 2000 2001 2002 2003 Round-wood 119761 61482 84750 91215 130290 93216 Sawn-wood 28180 15323 19392 29600 29248 29928 Parquet 16394 6446 9269 3937 3715 2920 Plywood 662 661 764 664 720 82 Veneer 2792 992 826 913 1130 15 Posts 8570 3219 1028 5006 3570 Source Relatho EstatisticoAnual, DirecloNacional de Florestase Fauna Bravia. 311, The export bano f first-class species roundwood aimed at increasingvalue added inthe forestry sector. With regardto forest fees, forest regulatorsgranted a 40 percent fee reduction for all logs that are processed domestically. Exports are a tiny fraction o f total production, having dramatically declined to less than 1percent o f the total output, also as a consequence o f the export ban. Inversely, processed wood has rapidly increased to the level of 3 percent. Altogether, wood exports generate a value o f only US$ 30 millionannually. 3 12. As i s shown inTable 34, forest license fees inMozambique differentiate among species. This facilitates the right rent capture. Between 2002 and 2005, the authorities have radically increased annual license fees. ''@relatedsubjecti 149 Its growthpotential would be evenhigher ifplantations with exotic, fast growing trees were takenup. A s that of fuelwood and charcoalproduction. For brevity, this issue i s not addressed. 90 CHAPTER5:NATURALRESOURCES Table 34. Licensefee structure per cubic meter, 1998-2005 (US$) Wood Class 1998 2003 2005 Precious 8.5 41.8 83.7 1'' class 5.2 10.5 20.1 2"d class 3.6 6.3 12.6 3 1class ~ 2.4 4.2 8.4 4'h class 1.6 2.1 4.2 Estere 31dclass 4.0 3.1 6.3 Estere 4" class 2.4 2.1 4.2 Charcoal 0.3 0.2 0.4 Others 2.0 2.1 0.2 Note: All US$values converted at annual average exchange rates 313. As can be seen from Table 35, fee revenues increased five-fold with the last fee revision, from Mt 11.9 billion in2002 to Mt 54 billion (US$2.26 million) in2003, and now account for 0.4 percent o f total government revenue. It appears that at this level the current fee scheme i s succeeding incapturing forest rents: the underlying background paper (see Margulis 2005) developed a methodology to estimate rent levels. Based on broadly accepted assumptions, it seems that the current average license fee for first class wood i s within the expected range for scarcity rent values, and that the expected doubling in2005 could be seen as the upper limit for an efficient rent pricing. Table 35. License fee revenues, 1998-2003 I Year Revenue (Meticais x lo6) Revenue (US$) 1998 3,377 270,000 1999 3,787 300,000 2000 7,47 1 530,000 2001 11,855 580,000 2002 11,856 500,000 2003 54,077 2,260,000 Source: R e l a t h o Estatistico Annual. Direccao Nacional de Florestas e Fauna Bravia 314, Managementpolicies, implementation and governance. The challenge o f the forestry sector inMozambique i s to find the optimal set o f incentives to promote forestry without exhausting the forest resources. The government needs to capture the forest rents associated with timber exploitation without creating inefficient barriers to output expansion or promotingtechnical inefficiencies, while at the same time protecting the poor and preserving the ecological services. The 1999 Mozambique Forest and Wildlife L a w was designed with this challenge inmind, and relies on modem forest principles and instruments. 315, Volume licensefees and rent pricing. Recent surveys inthe forestry sector in Mozambique indicate a highdegree o f technical inefficiency throughout the production chain, suggesting that the current rent pricing based on licensing fees may not be enough to create compatible incentives. The forest law states that both forest concessions and simple licenses can be issued independently o f forest productivity. Logging companies 91 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM thus tend to maximize logging volumes while benefiting from the advantages o f simple licenses, bringing the forest value down. A disadvantage with the current system o f fees ("volume license fee") i s that it differentiates by wood type, which i s appropriate, but does not differentiate by locality. This gives undue incentives to harvest wood near populous areas, and discourages harvesting in distant areas. This is an inefficient arrangement, causing under-use inremote areas where there are no ecological problems, and threatening deforestationinareas o f easy access.151 152 316. The economic literature advocates biddingas the most appropriate rent pricing mechanism, as operators reveal their maximum willingness to pay according to their expected flow o f revenues and costs, allowing resource owners to capture all rent. It also discourages rent-seeking behavior observed inadministrative licensing authorizations. Onthe other hand, biddingmechanisms require some basic conditions: zoning o f concession areas with reliable inventories; participation o f a sufficient number o f operators to make the process competitive; and an efficient and credible system o f monitoring and sanction fees, none o f which are fully met inMozambique. Hence it would be prudent to introduce the mechanism only gradually. 3 17. Incentivesfor woodprocessing. As mentioned above, to increase value added, the authorities banned exports o f first-class species roundwood, and granted a 40 percent fee reduction for domestically processed logs, and a 75 percent reduction for two species that account for 53 percent o f total output. Banning log exports may not constitute an appropriate measure to promote domestic processing as it tends to stimulate rent seeking and illegal exports. Some studies have suggested that the ban o f log exports in tropical countries tends to depress log values and incomes by more than the value added in processingand later export. Furthermore, the existingplants are as yet unable to process all the wood available, and not surprisingly the total amount o f wood produced has fallen. More effective, and fiscally advantageous, would be a carefully graduated combination o f time-bound tax and licensing advantages, a time-bound export tax at a light rate, and infrastructure incentives. 318. Inthe early stages o f setting up o fthe industry, stumpage fee reductions can make a significant difference, particularly in a country like Mozambique, with serious domestic credit restrictions, highinterest rates, and limited skilled labor and business capacity. A phase-out program to gradually eliminate such fee reductions in five years may be appropriate, considering the investment cycle o f the sector. Interms o f total tax revenues, such temporary fee reduction may end up increasing total tax collection inthe long rundue to the expansion o f industrial activities, employment and good circulation. 3 19. Currently, licensing can be granted either for concessions or for simple operators. Simple operators are only required to present a simplifiedmanagement plan, with information on area, volume and species. Managerial practices and exploitation planning are not required. Simple licenses are granted only to nationals and up to the volume o f lS1 Obviously the "first-best" arrangement would be a bidding scheme, as is discussedbelow. But this i s a long way off inMozambique. Inthe meanwhile, fee differentiation by locality is advocatedas a "second- best" solution for limiting environmental damage. lS2commentatorarguedthatinsteadoflocalitythefeeshouldtakeintoaccounttheproductivity of One forests, andthat the fee shouldbe higher for cutting inthe productive forests in, say, Nampula. But this would have the effect of accentuatingdeforestationinareas of easy access. 92 CHAPTER5: NATURAL RESOURCES 500 m3per year. The contribution o fwood supply from simple licenses i s expectedto decline with the reduction o f the fee waiver. The highnumber o f simple licensees makes monitoring too costly, reducing performance, and inducing operators to use up the forest with minimal ecological and efficiency concerns. 320. Concessions, on the other hand, are required to have a detailed management plan. Butthese require muchtechnical expertise and investments ininventory data, and with the underdeveloped domestic capital market, such plans are only affordable to highly capitalized foreign companies, so that only four management plans have been approved out o f 45 authorized concessions inthe period 2003-04. 321. Inthe lightofthese problems, animmediatetransitionto abiddingregimeis impossible. It i s hence proposedthat a declining fee reduction arrangement (phased out over five years) be used instead. 322. The tax mechanisms have promising potential for increasedrevenues, possibly US$ 1.5 million: P The differentiation o f license fees canbe improved-by adding differentiationby locality -but without any overall increase inrevenues, given the doubling o f fees in2005. In2003, revenue increasedby 500percentwhereas average fee went up by a higherproportion. We conservatively assume that the proposed fee rates made an additional 50 percent increase in2003 revenue value, that is, US$ 1.15 mi1lion. P The existing fee reduction onprocessedlogis on average 60 percent. Assuminga 30 percent share o f processed wood intotal revenue, and assuming processing losses o f 50 percent, and assuming exempted logs are about 60,000 m3,and assuming an average fee for first class wood o f US$10, we estimate the total foregone revenue at US$360,000. 323. Benefits to communities. Community benefits are the central issue inthe current forest regulation inMozambique. Two important instruments inthe Forestry Regulation take communities into account: Article 102 states that 20 percent o f the license fee revenue should be devoted to communities living inthe areas where resources are extracted; and Article 36 requires concession proponents to conduct a "public audience" before an authorization is conceded. Neither clause i s very effective. The 20 percent o f the license fee ends up in the general budget and i s rarely passed on to the community.'53 Results o f the "public audiences" have been mixed, varying from a simple consultation with little given up to communities, to a very complex bargaining process. Inthe latter case operators usually end up offering significant side payments either incash or in social investments (schools, roads, wells, etc). It would be well for the Government to investigate indetail the success o f the consultationrequirement inprotecting community rights. 324. Recommelzdations. Enormous regulatory efforts in forestry have already been made inMozambique, not only interms o f legislation, but also in implementation and analytical work. However, more i s needed inorder to benefit fully from the growth and 153 DiplomaMinisterialNo. 9312005 (May 4) provides for mechanisms o f passing on o f these funds. Time will tell whether it effectivelydoes that. 93 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM poverty reduction potential o f the sector. Growth, efficiency and social benefits can only keep increasing andpreserving the country's natural basis if logging prices are correctly set. Although administrative forest rent pricing must be immediately improved to accomplish that, the medium-term goal i s the introduction o f a biddinglicense regime. To accomplish this: 9 Develop aprogram for implementingazoningsystemofloggingareasbasedon detailed inventories, optimal exploitation sizes, and indicators o f social conflict. This is the most important first step to set the basis for introducing bidding license mechanisms. P Redesignthe volume licensefee to take into account locationvariables that also affect rent. Moreover, in accounting for efficiency levels, it should create incentives for better managerial and technical practices. It should additionally set periodical revisions o f rent values. P Phasingout the fee reduction. The current 40 percent fee reduction for processed logs i s not dynamically efficient. This fee reduction must be seen as temporary subsidy and a phased out program must be designed and fully publicized among beneficiaries. P Either establish an independent monitoring/auditingbody or strengthen the existing monitoring arrangement, so as to assure non-routine inspection o f the forest operations as an initial step to improve the transparency o f attribution and management o f forest concessions and simple licenses. P Biddingmechanisms. To prepare for these, the Forest Service should ensure that the basic premises o f forest concessions are met, particularly timber stock and forest productivity. The Forest Service needs to increase the quality o f information o f the concession areas and reduce the asymmetry o f information, which i s presently infavor o f the concessionaire. 325. To encourage the creation o f additional value added inthe sector, projects could be set up for the creation o f industrialzones for wood processing facilities combined with fiscal incentives and matching grants for staff training. 326. Finally, the forestry sector is also important as a contributor to the livelihood o f tens o f thousands o f rural dwellers. Two recommendations follow from the discussion above: 9 The present arrangement ofissuingsimple licenses for 500 cubic metersof cutting annually should be maintained untilsuch time as a fully-fledged bidding system can be introduced. 9 The approval andimplementation ofrevenue sharingmechanisms for community benefit should be accelerated, with the objective o f complying with the social objective o f the forest regulation and ensuring a more effective monitoring system. Inorder to limit the erosion of communities' land and forestry rights, the effectiveness o f the consultationrequirement for concessioning should be evaluated. 94 CHAPTER 5: NATURAL RESOURCES F. MINING 327. Mozambique has favorable geology for mining activities. The sector comprises (mostly informal) small scale and manual operations, extracting gold and precious and semi-precious stones. Miningaccounts for just under half a percent of GDP. The value o f mineralproduction on an annual basis i s presented inTable 36. Table 36. Value of annualmineralproduction(US$) Mineral 1999 2000 2001 2002 2003* Sand 898,620 1,394,052 2,387,439 4,116,096 Clay 2,011,220 2,051,559 2,730,773 3,255,720 Bauxite 567,576 585,367 618,631 656,568 849,117 Coal 94,306 177,265 303,596 478,632 404,162 Limestone 5,855,900 7,292,300 13,012,323 13,483,720 Gold 186,556 224,065 218,310 167,702 623,176 Tantalite 625,000 675,000 1,172,500 4,717,375 Buildingstones 2,211,324 3,612,845 3,022,295 4,774,395 4.443,008 Others Total 7,159,836 15,861,359 16,552,082 26,293,23 1 32,789,559 jource : DNM c 2003 Droduction up to September ~ 328. Large scale miningemployment i s some 3,000; artisanal or small scale mining employment exceeds 50,000, or 100,000 if seasonal and occasionalwork i s included. Industrial production i s limitedto tantalite andbuilding materials. 329. Fiscalrevenues from the sector rose from US$ 0.8 millionin 1997 to US$ 3-5 million annually at present. 2005 saw a sharp increase inrevenues when a sum o f US$120millionwas paidbythe BrazilianfirmCompanhiado Vale do RioDoce (CVRD) to the government for exploration rights inthe Moatize area, with a view to preparing a project for coal mining. Since small scale production o f gold and bauxite restarted inManica in 1990, exports by the sector rose from US$4.5 million in 1992 to US$32 millionin2003. Inaddition, itis estimated that US$ 10millionofgold andUS$ 30 million o f semi-precious stones are exported through informal channels annually, which represent some US$ 1millionin forgone fiscal revenues at current royalty levels. 330. The number o f formal sector firms doubled between 2001 and 2003, and the number o f foreign firms rose from 30 to 77. Two heavy mineral sands projects are at an advanced stage o f development. (i) The Moma project has a total capital cost o f US$220 million. At full capacity (expected in2006), the project should produce 625,000 tones per year of ilmenite,34,000 tones of zircon, and 14,000 tones o frutile, generating annual export revenues o f about USS 80 million and employing directly 425 people. (ii) The Limpopo Corridor Sands will to beginproduction in2008. The first phase o f the project 154 Instituto Nacional de Estatistica, "Industria extrac$io de minerais", online table, M a y 2005, 95 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM is expectedto produce 375,000 annual tones of titanium slag, with a capital cost of about US$ 500 million. Annual sales are estimated at about US$300 million, with about 900 newjobs created. A second phase i s expected to increase capacity to 1million tones per year o f titanium slag.'55 331. The sharp increase in sector activity has been associated with the improvements in the law and regulation of mining, and also the introduction of an innovative and transparent MiningCadastre.'56 Exploration expenditures runto some US$15-25 million annually. 332. The recent trends are encouraging, but Mozambique's miningoutput is still well below that o f countries with similar geological potential. Annual exploration expenditures could easily rise to US$ 50-100 million per year. Given the advances in sector reform achieved recently, the increase inthe quantity and quality o f geological information available, and the improvements inthe quality of the services delivered by the relevant government institutions, Mozambique's mineral exploitation i s likely to have a prolonged period o f rapid growth. With the two heavy sands projects alluded to above, and the recent start o f the rehabilitation o f the Moatize coal fields, export revenues could reach US$500 million by 2010 and US$700millionby 2015. 333. Interms ofpoverty reduction, duringand after the conflict period the extraction o f gold and gemstones has been serving as an economic buffer, being one o f the few activities providing immediate cash income for the poorer Mozambicans. For the vast majority o f them, miningi s a part-time activity undertakeninparallel with subsistence agriculture, being an intimate part o f rural societies, traditional village structures and their economies. Miningsector expansion has been associated with a reduced rate o f poverty. While the poverty headcount (as measured by adult equivalent consumption) declined nationwide from 69 to 54 percent between 1997 and 2003, the miningsector showed the swiftest decline, from 68 to 44 percent. Table37. Poverty rates by sector of employment of the householdhead Sector 1996/7 200213 Mining 67.7 43.5 Manufacturing 61.9 52.1 Construction 63.7 59.1 Transport 53.9 35.2 Trade 54.3 42.1 Services 52.3 50.1 Education 53.0 27.1 Health 61.0 43.0 Public Administration 49.4 35.1 Agriculture 72.5 ___ ___ 58.1 Source IAF 199617and 200213, and Fox ef al. (2005). Poverty rate' by per adult equivalentconsumption. All individuals in householdare assigned to the sector of employment `55 See more on the mega-projects in Chapter 1 (page 23). `56 This innovative reformto the Cadastre i s an excellent example o f growth-inducing institutional change. See below inparagraph 336, page 97, and paragraph371, page 109. 96 CHAPTER5: NATURAL RESOURCES of the householdhead. Ifthe headis not employedthey are assignedto the sector of employment of the oldest adult. If nobody works (less than 5% of all cases) they are assignedto agriculture. See Annex Table 22 (p. 152) for further detail. Part o f the explanation for this finding i s the growth o f artisanal mining, which i s the only cash-generating activity in some rural areas. 334. Despite the obvious potential benefits from the activity, the associated costs and benefits may accrue unevenly to different groups. Miningcan exacerbate poverty directly by damaging the environment on which subsistence economies dependon, by creating new social and economic problems, while communities are often not given full opportunity to participate indiscussions on proposed projects. The outcome will often dependon policy and institutional frameworks inplace, as well as the government's political commitment to support often dispersed and isolated, less informed and less empowered groups. 335. Managementpolicies, implementation and governance. Mineral resources are the property of the State. The Ministryof MineralResources and Energy(MIREME)has administrative and regulatory responsibility for minerals, with the NationalDirectorate o f Minesbeingresponsible for licensing procedures. The main goal o f government policy for mining(Resolution 4/98 o f the Counsel o f Ministers) i s to establish a socioeconomic environment favorable to the development o f the national private sector and to attract foreign investment. The new policy shifted the role o f the State from producer to promoter andregulator, leaving the operational and implementation role to the private sector. Reforms carried out since the publicationo f the new mineralpolicy have included: 9 Revisionof the miningcode anditsregulations, attractingprivate investment into geological exploration and mining, and incorporating environmental provisions in coordination with MICOA; P privatizationofthe state-owned enterprises, limiting Stateparticipation to ventures with prior State investment; > capacity development, training and creation o f mechanisms for efficient supervision and monitoringo f miningactivities; 9 support to small scale andartisanal mining,trying to bringoperators intothe formal sector through the provision o f geological services, technical support, and information. 336. The most significant contribution o f the MiningCode to improve governance in the management o fmineralresources was the establishment, in2003, o f a non- discretionary and transparent system to manage miningpermits -the MiningCadastre. The Cadastre enforces a simple set o fregulations to promote a level playing field in access to mineral resources by the private sector. Basedon a "first come, first served" principle, the Cadastre grants mininglicenses based on simple conditions and the payment of a fixed miningadministration fee. Holding o f permits depends exclusively onthe payment o f surface rental fees, which are proportional to the area o f the title. As a result the number of new applications for miningtitles has risen (2 13 in 2004, compared to 140 in2003, and 110 in 2001). So has the number o f valid licenses, (325 in2001, 550 97 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM in2003). The averagetimerequiredto obtainaminingtitle has droppedfrom 120days in1999 to 50 days in2003. 337. Byreducing the discretionarypower o f Government, increasingtransparency, and improving security to potential investors, the new MiningCode has provided the adequate legislative and regulatory framework for promoting, regulating and monitoring large industrial activities. Its success should clearly serve as an inspiration for the management o f other natural resources. 338. The system o f mining taxation system consists o f the general income tax code (Impost0sobre o rendimento daspessoas colectivas -IRPC), plus the royalty (10-12 percent for diamonds, 3-8 percent for minerals), and the surface tax associated with the holding o f permits. Miningcompanies are also subject to local government taxes. 339. Inthe case o f"mega-projects" the fiscalprovisions ofthe MiningCode usually apply to the extraction activities. Inorder to attract foreign investment, the Government has resorted to one-on-one special arrangements, but this has its own problems: Mozambique i s not getting its fair share o f the resource rents, and monitoring compliance with tailor-made tax regimes is a prohibitively bigtask. Inaddition, beneficiation operations have benefited from the Industrial Free Trade Zone incentives, which establish a ceiling for the tax burden o f 1percent o f net tumover. The latter provision creates an additional problem for the authorities because free market prices do not exist for most o f the raw materials involved, leaving tax collection vulnerable to "transfer pricing". 340. The system i s unattractive to private investors because (i) the highroyalties can place a large burden on companies duringmineral market downturns, (ii) the IRPC i s not suited to resource rent industries, (iii) the IRPC does not permit accumulation and carry forward o f expenditures duringthe exploration and development phases, or accelerated depreciation for these exploration and development expenditures. 341. For mininginvestments, the Cddigo dos Beneficios Fiscais (Code o f Fiscal Incentives, 2002) provides for a tax reduction o f 25 percent for investments valued at over $500,000, up to 2010 (Article 42). Inaddition, for all investmentswhether mining or not, the Code provides an investment tax credit o f 5 percent o f the investment realized, for five years (Article 15). For a period o f time, these measures will reduce some o f the monitoring costs, but the lack o f adaptation to the s ecificity o f miningventures remains (lack o f carry-forward, accelerated depreciation)."` Remedies are discussed below in paragraph347 (p. 99). 342. Hence the arrangements for mining firms need to be revised from the roots up, to make them globally competitive, to secure a fair share for the exchequer, and to ensure neutrality across sectors such that investment decisions are not affected by differential taxes but by differences inpotential profits and risks. 343. Poverty reduction. The Government o f Mozambique has recognized inits Poverty Reduction Strategy (PARPA)that small scale and artisanal mininghave a role to play inlinking sector policies with national poverty reduction goals. Because o f its intensive use o f labor, it provides one o f the few forms o f self-employment and income generation in the rural areas. 157 See www.tralac.orgiscriptsicontent.php'?id=lOSSfor further information on the Code. 98 CHAPTER5: NATURAL RESOURCES 344. The miningsector can contribute to poverty reductiondirectly and indirectly. The direct links are mainly through small scale artisanal miningwhich i s traditionally conducted by families and entire communities off the agricultural season. It createsjob opportunities for local people, encourages localbusinesses, and contributes to the provisiono f vital infrastructure for remote communities that do influence development positively, such as roads, education andhealth care facilities. According to the household census o f 1997, miningi s an important source o f diversification o f non-farmrural incomes inthe countryside. 345. The indirect links are mainly through supply-chain mechanisms inlarge mining projects. Increasingly, scores o f micro and small enterprises that are owned primarily by Mozambicans are providing services for the mega-projects (see paragraph 77, page 24, for further detail). The other main indirect link i s through government revenues; thus the challenge i s to improve the efficiency o f expenditures, as has been argued inChapter 1 (para. 33, page 11). 346. The impact of the miningsector inalleviating poverty canbe enhanced. Despite the government's progress with the new MiningCode facilitating access to minerals, rural farmers still face procedural difficulties when trying to secure nationally recognized legal rights to the minerals they extract. When a company is granted legal rights by the MiningCadastre, local communities maybe forcedto sell all the minerals output they extract from their traditional lands or face the risk o f losing access to land that holds cultural significance to them and may be their chiefmeans o f subsistence. The government strategy to cope with this situation i s to decentralize the miningcadastre's services bringingthem closer to the final clients, and to provide extension services to artisanal miners, facilitating their formalization. Inthe case o f large mechanized investments, the regulatory framework clearly recognizes that local communities need recognition inthe form o f consultation processes, as well as fair benefit sharing and trusted grievance mechanisms. 347. Recommendations. Inorder to maintain the mining sector's leading role in growth and inpoverty reduction, there is a pendingagenda o f reform. (a) Improving governance in revenuemanagement. Although the Government recognizes that a special tax treatment i s needed to accommodate the specificities o f mining, this should be done in such a manner that the treatment i s automatic for all miningfirms, rather than provided on aproject specific basis or inthe form of special incentives. The following principles should be observed inimproving the mining taxation package for Mozambique: P Taxes shouldbe consistent with those of other sectors ofthe economy. All sectors o f the economy shouldbe subject to the same - or very similar -taxation, for investmentsto be made efficiently. Such tax neutrality across sectors means that investment and production decisions should not be affected by differential taxes or incentives, but by sector differences inpotential profits and risks. k The fiscal regimeshouldbeprofit-basedratherthanproduction-based, avoiding excessive royalties, which have to be paid independently o f profitability. k Taxes shouldbestable over the longterm: the decisionto develop amineentails huge commitments for the life o f a project with long return on investment periods; 99 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM i t i s therefore important that the country's mineral taxation regime be stable. Miningcompanies needassurances that their projectswill not be affectedby adverse changes on the mininglegislation and related sectors, such as environmental legislation, federal taxation, and trade and foreign exchange policies. Therefore, a Stabilization Agreement that provides explicit, standardized, and non-negotiable guarantees o f stable treatment with respect to these aspects should be core parts o f Mozambique's mining investment agreement. 9 The tax regime shouldbe adaptedto reflect some o fthe sector's major characteristics: (i)miningis a highrisk, capital intensive industryandtherefore the fiscal regime should allow the miningcompany to recover its capital investment before paying substantial amounts o f taxes; (ii) income taxes should recognize that miningi s a cyclical industrymarked by fluctuating market prices, and provide loss carry-forward rules and the carry-back o f mine closure costs; and (iii) fiscalregimeshouldtakeintoaccountthefactthatmineralsarenon the renewable resources and therefore allow for all exploration and development costs to be expensed or amortized when incurred. (b) Mainstreamingartisanalmining. Because o f lack o f institutional capacity o f the administration, most o f the small-scale and artisanal mining i s being conducted outside the formal channels, with insufficient linkages to the rest o f the economy, damage to environment, social conflicts, and loss o f fiscal revenues. The government needs to implement policies to better integrate these individuals and small firms with the rest o f the economy. Inconcrete terms, this calls for: P capacity buildingo fthe Provincialminingadministration, with the establishment o f a better alignment between core central public sector functions and the implementation o f their mandates at the Provincial level; P definition o feffective means for community empowerment andparticipation, including the establishment o f public-private partnerships with responsiblemining companies willing to invest part o f their profits in human resources, and social and physical infrastructures; P provision o fdirect educationand training on health (including HIV/AIDS awareness), safety and environmental impacts to miners, local community members and local governments. This can be effective inhelpingminers minimize the hazards o f small-scale mining, including damage to the mining area and surrounding communities; and 9 at themunicipal level, thebiggestchallengewill beto increasethe tax revenues o f the municipalities through the decentralization o f tax collection mechanisms. The on-going reform o f the miningfiscal regime (incoordination with the Ministry o f Finance) could increase fiscal revenues inresource rich municipalities, and channel such revenues directly to them. 100 CHAPTER 5: NATURAL RESOURCES G. CONCLUSIONS 348. The main findings from this chapter are succintly summarized inTable 38. Major growth opportunities - given appropriate reforms - are presentedby the resources o f rural land, water and mining. Land and water reforms can help promote agriculture, which i s the second most important motor o f growth and the most important sector for poverty reduction. Mining, particularly but not exclusively through mega-projects, has tremendous potential for expansion. Table 38. Naturalresources: summary of findings and recommendations Natura1 Major growth Major Chief policyrecommendation resource opportunities? importancefor subsistence? Ruralland Yes -medium & large Yes Enable land to automatically farming accompany sales o f buildings and improvements Water Yes - irrigated agricul- Yes - rural water Facilitate small-scale irrigation; ture, and reducing costs supply, urban water vastly expand rural and urban o f droughts & floods supply household water supply Fisheries No Yes Increase the tax take, andpursue transparency in allocation of quotas Forestry Yes. Output can double Yes Transparent allocation but the sector as a whole i s small. Further opportunity with plantations. Mining Yes - mega-projects No (% of house- Maintain transparent Mining Code holds i s small) and Cadastre 349. The picture is very different when thinking o f the importance for subsistence o f large numberso fMozambicans. Land, water, fisheries and forestry are o f key importance for subsistence o f millions o f people. It i s important that their access to livelihoods not be upset as the country's developmentproceeds. For example: reforms are clearly neededin fisheries, mainlyby way o f increasing the tax take and pursuing transparency in quota allocation; but it i s important that the free access enjoyed by the subsistence fisheries sector not be eroded as some 90,000 people depend on it. 350. The "short list" o f reforms proposedinTable 38 i s ambitious but feasible. Most important, for its capacity to reduce poverty, i s to vastly increase rural and urban water supply.Rural land regulations needto be changed such that sales ofbuildingsand improvements are automatically accompanied by the landconcession, without further official ministrations. This will facilitate landtransactions and increase the sense o f land security, thereby encouraging investment. Inmining,the most important action i s simply to continue the current transparent system o f administration via the Mining Code and Cadastre as this has already demonstrated its value inattracting investment. 101 CHAPTER 6. GROWTHAND POVERTY OUTLOOK, AND KEY RECOMMENDATIONS 351, This concluding chapter seeks to peer into the future to see what growth and poverty outcomes are likely to be, and summarizes the key recommendations from the preceding substantive chapters. A. GROWTH THELONGTERM IN 352. H o w much growth can Mozambique expect inthe long term? This chapter attempts an educated guess at likely growth rates from three sources: (a) a comparison with Mozambique's neighbors, and a general equilibrium model, and (b) an intercountry comparison using the "growth regression" literature, and (c) and examination o f the institutional requirements for higher growth rates. It then briefly examines the growth prospects inkey sectors. Finally it draws out the likely implications for the evolution o f poverty over the next decade. (a) A regionalcomparisonand a generalequilibriummodel 353. The Government's predictions o f roughly 7 percent GDP growth in2005 and 2006 appear reasonable and consistent with growth rates inthe recent past. The following discussion i s focused, therefore, on the medium and long term, from 2007 onwards. 354. Growth inthe case of Mozambique's neighbors was a more modest 3.4 percent on average between 1995 and 2002 (Table 39). The reasons for Mozambique's superior 8-9 percent growth record inthe period have been examined: inpost-conflict situations "growth spurts" are frequent as agricultural yields catch up and traditional markets are recovered; investment from the region found Mozambique a useful haven; Mozambique has managed to attract a number o f mega-projects; Zimbabwe expelled its farmers who built up Mozambique's tobacco industry;Mozambique has an excellent coastline and is experiencing a tourist boom; andMozambique has attracted more aid than most o f its neighbors. 355. Inthe long term some of these elements will cease to count for Mozambique: the post-conflict catch-up will tail off, the Zimbabwean farmers are a one-time gain anyway, and aid disbursements may gradually decline to the regional average.'58 Inthe long run, Mozambique's advantage inthe region will consist o f its ability to continue to attract mega-projects, and its tourist potential. Hence we may still expect Mozambique's growth to be better than the region's, though by a lesser margin thanhas been observed inthe past. And since the mega-projects have historically had a limited impact on private 15* Collier and Hoeffler (2002) find that the absorptive capacity for aid is no higher than "normal" in the first three years o fpeace, but that it rises to double its normal level thereafter, lasting untilthe end o f the decade, after which it tails o f f to the "normal" level. MOZAMBIQUE COLTTRYECONOMIC MEMORANDUM consumption, the main factor making for a faster growth rate ofprivate consumption will be the unexploited tourist potential. Hence for this Memorandum the long-term growth rate used for projections should certainly exceed 3.4 percent, but perhaps not by a wide margin. Table 39. GDP growth rates of Mozambique'sneighbors 1995-2002 Ethiopia 5.0 Kenya 1.4 Lesotho 2.1 Madagascar 2.8 Malawi 2.2 Tanzania 4.6 Uganda 6.1 Zambia 2.7 - Mean 3.4 Median 2.8 - Mozambique 8.6 Source: World Bank, SIMA (Statistical Information, Management and Analysis). GDP is in constant local currency units. The growth rate is the least sauares mowthrate as described in Table 7. 356. Another way o f azing at the crystal ball i s to start with the general equilibrium model referred to a b 0 v 3 ~which in unadjustedform delivers an average long-term ~ growth rate o f 7 percent. This growth rate reflects the amounts o f investment and saving inthe recent past (including direct foreign investment and aid), together withtheir effectiveness intranslating investment into output expansion. But this growth rate has to be temperedby two factors: natural disasters and HIVIAIDS. 357. Repeated major meteorological disasters are a virtual certainty: from 1965 to 2001, 14 major floods, nine major droughts and four major disasters from typhoon landfalls occurred. Most were associated with outbreaks o f cholera and diarrheal diseases (Kostermans, 2000, p. 2). Mozambique's proneness to natural disasters i s well illustrated inFigure 6. Major events with substantial impacts on growth have occurred at the rate o f roughly one in five years. 159 Paragraph 84 (p. 26) and Benito-Spinetto and Moll (2005, Appendix). 104 CHAPTER6: GROWTH AND POVERTY OUTLOOK Figure 6. Natural disasters and GDP growth, 1980-2004160 358. RegressingGDP growth 1981-2004 on current-year investment, a dummy for the war years and a natural disaster dummy yields lOO*AY/Y =10.2+ 2.1 (IIGDP) - 5.6 Dnar-dis - 9.2 D,,, , R2=0.43 (14) (2.9) (3.7) where the standard errors are inparentheses. The regression shouldbetaken with the utmost caution.*61Nevertheless the finding that war andnatural disasters are bad for growth i s unlikelyto be overturned. The regression suggests that growth i s cut by 5.6 percentage points when a major disaster strikes. At the rate o f one major disaster every five years, on average GDP growth is reducedby about 1percent annually.'62 359. Growth and HIV/AIDS. Mozambique has not yet won the battle against HIV/AIDS. Mozambique i s one o f the African countries hit hardest by the HIV epidemic, In 1992, the adult HIV prevalence rate was 3 percent'63. By 2002, 1.1 million Mozambican adults and children were living with HIV/AIDS, with an adult prevalence o f 13.8 percent. By 2004 the prevalence rate was 16.2 percent. The number o f AIDS- related deaths i s projectedto double by 2010. By 2010 life expectancy at birth i s expected to drop from 43 to 36 years, rather than increasing to 50 years as it would have inthe absence ofHIVIAIDS. Similarly, by 2010 infant mortalityis expectedto be at least 25 percent higher than it would have been inthe absence of HIV/AIDS. By 2010, an estimated 1.13 millionMozambican children will have lost one or both parents to AIDS. I 6 O Sources: GDP from INE and Ministry o f Finance, together with mission estimates; rainfall data from Rimma Dankova, private communication. The vertical lines do not (obviously) represent the only fluctuations inrainfall. The criterion used was that the disaster have national proportions. 16' The regression should be taken with the utmost caution because it lacks variables known to be associated with growth and the investment variable is statistically insignificant. Also, the coefficient o f determination i s 0.46, indicating that there i s much about growth that the regression approach i s incapable o f capturing. '62 The implications o f natural disasters for poverty have been referred to inChapter 5 (para. 262, p. 78) and have been covered ina contemporaneous study (World Bank, 2005b). WorldBank, 2004d, p. 16. 105 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM 360. The demand for health services will increase substantially as a result o f the HIV/AIDS pandemic. HIV/AIDS affects the health sector both by increasing the demand for health services and taking its toll on health personnel. Some o f these increasing costs are already being borne by the system inthe form o f treatments for opportunistic diseases; an estimated 30 percent o f hospital beds are occupied by HIV patients (Haacker, 2001). The likely cost o f treatment inthe future depends on the government's decisions on coverage and nature o f care. To provide palliative care and treatment o f opportunistic diseases to 20-30 percent o f the population with HIV would absorb halfo f total health expenditure (Haacker, 2001). The costs o f the loss o f professionals inthe health sector i s substantial. Training o f doctors and nurses will have to be expanded by at least 25 percent over the 2000-2010 periodjust to keep the number o f doctors and nurses constant (Haacker, 2001). 361, The government has recognizedthe threat HIV/AIDS poses and i s acting to prevent and mitigate its effects. Interventions currently underway or planned include those supported under the World Bank-financed HIV/AIDS Response Project and Treatment Acceleration Program, the Global Fund, the Clinton Fund, as well as several bilateral agencies. The challenge will be to ensure that all o f these programs are well coordinated, target the poor for prevention, care, and alleviation. The government should also ensure that the additional fundingdoes not cause distortions inthe health delivery system, including changes inthe demand for inpatient and outpatient services, consumption o f drugs, and health personnel workload. 362. Arndt (2003) uses a computable general equilibriummodeling approach to study the impact o f HIVIAIDS on human capital accumulation and economic growth in Mozambique. The impact i s estimated for three AIDS scenarios differing inthe assumptions concerning the impact o f HIV/AIDS on (1) productivity, (2) on population, labor and human capital accumulation and (3) on physical capital accumulation effects. H e finds that the Mozambican economy will be 14percent to 20 percent smaller in2010 on account o f AIDS. Per capita GDP growth will be reduced by 0.3 percent per annum at the start o f the period, and 1.O percent lower per annum by the end o f the period. 363. Putting these elements together, Mozambique's long-term growth i s likely to be about 5 percent (= 7 percent from the model, less 1percent per year for HIV-AIDS, less another 1percent per year to take account o f meteorological risk). This i s on the assumption o f the institutional environment remaining much as it i s at present. N o w we examine what could happen ifdetermined efforts were made to strive for better growth and poverty reduction outcomes. (b) An intercountrycomparisonusingthe "growthregression" literature 364. Simulations were conducted usingthe "growth regression'' literature, comparing Mozambique's policy performance with that of some o f the best performers in the developing world, including East Asian c ~ u n t r i e s . 'The method was to identify some ~ ~ carefully executed regression studies, and plug altemately Mozambican data and "best performer" data into the equations, to see what growth increment could be obtained by improvingMozambique's performance to that o f the best. The results showed that 164 See Benito-Spinetto and Moll (2005) for details o f the methodology and results. 106 CHAPTER 6: GROWTH AND POVERTY OUTLOOK Mozambique could improve its growth performance markedly by raising its performance to East Asian levels. Modest gains - 0.5 to 1percentage points per annum - could be made by P avoidingbankingcrises P improvinggovernance P reducing governmentnonproductive spending. Larger gains, of as much as 3 percentage points o f growth per annum - over time - would stem from P improvinginfrastructure. Mozambique hadonly about 4 telephones per 1000 inhabitants during the period inthe late 1 9 9 0 while Malaysia had 190. ~ ~ P raising access to secondary education. InMozambique the net secondary school enrollment rate averaged only 8 percent in the late 1990s. InMalaysia the figure was 80 percent. Obviously making up gaps o f this magnitude will take many years. Hence in the medium term the growth "kick" from further improvements ininfrastructure and secondary education is likely to be much smaller. 365. None o f these pointers shouldbe taken inisolation from one another or in isolation from the overall policy stance o f the government. For instance, improving telephones alone without simultaneously improving other aspects o f infrastructure such as roads, and further facilitating the business and investment environment, i s not likely to turnupthe GDP growthimprovements suggestedbythe regression models. But suppose Mozambique does make reforms on a broad front, and inparticular makes improvements inall five ofthe areaspointedupbythe regression models. Then it is not impossible that from each might arise an additional 0.5 percentage points o f GDP growth per year, for a total increment of, say, 2.5 percentage points. Again, this number should not be taken literally, as the phenomenon o f growth is still imperfectly understood. Nevertheless, the qualitative conclusion from these experiments i s clear: Mozambique can considerably improve its growth performance by undertaking reforms inthe investment environment and government performance. (c) Futuregrowth and institutions 366. The importance o f institutions for growth and poverty reduction was demonstrated by many examples inthe 2002 WorldDevelopment Report: Building Institutionsfor Markets. Rodrik (2003) has pointed out that many countries have experienced "growth spurts'7165 o f 10 years or more, frequently inresponse to a relatively narrow range of reforms (exchange rates, tariffs, competition), but that almost all o f these growth spurts have eventually petered out. The factors that make for a growth spurt - such as stability after a conflict, a competitive exchange rate, the accession o f a pro- growth government - are not necessarily the same as the factors that make for long-term sustained growth. For sustained growth a broader set o f institutional changes is called A "growth spurt" is defined as an increase in an economy's per capita GDP growth o f 2.5 percentage points or more, relative to the previous five years, that i s sustained over at least 10 years. 107 MOZAMBIQUECOUNTRY ECONOMICMEMORANDUM for, so as to achieve property rights, market-orientedincentives, sound money, and fiscal solvency. With this inmind, it i s not clear whether Mozambique's amazing growth spurt will be repeated inthe future. 367, The country would do well to seek for higher growth by introducing the institutional reforms that would shore up its long-term growth potential, so as to increase the system's ability to respond to shocks (e.g. floods, droughts) andto enable it to absorb technology. So far Mozambique has experienced a remarkable growth spurt despite a highlyunpropitious institutional environment inmany respects (see Chapter 4): a weak judicial system,'66 antiquated business legislation, an inability to enforce contracts or retrieve collateral, outdated labor laws, inflexible labor markets, the lack o f a formalized accounting profession, a troubled financial sector, unclear land rights, a moribund fanner cooperative system discredited by the socialist experiment, nascent but still undeveloped "civil society" institutions, low levels of capacity inmany government departments, a "red tape burden" for businesses and individuals so heavy that Mozambique scores near the poorest on several indexes. Inaddition, the infrastructural environment for business was unattractive, although subsequent improvements are raising Mozambique gradually to SADC levels -roads were poor everywhere except inthe major cities, water supply was intermittent, and electricity supplywas subject to frequent blackouts. The business- related institutional aspects were covered well inthe Investment Climate Assessment (World Bank, 2003a), and are dealt with inChapter 4. A proposed sequencing o f private sector-oriented institutional reforms was presented inTable 28 (p. 68), and hence i s not repeated here. 368. Onthe other hand, there were compensatinginstitutional factors: a fairly unified government with a firm commitment to poverty r e d ~ c t i o n ' ~an ~ ,ability to embark on (often painful) reforms as needed, the absence o f armed conflict, an active press, an unfettered information technology market permitting free use o f intemet capabilities, steady progress inlowering intemational trade barriers, free entry into agricultural marketing, easy access to visitors' and business visas, a functioning price information system for agricultural products nationwide, and a customary land tenure system that flexibly and nonviolently enabled the retum o f hundreds o f thousands o f persons displaced by the war. 369. As was discussed inChapter 1, one o f the reasons that Mozambique's growth record has been good despite its unpromising business environment i s that some o f the growth observed has been o f the "catch-up" variety. For instance, it was not difficult for banksto expandcredit to long-time clients, despite thejudicial problems with collateral. Bythe same token, however, the next stage o f growthwill not be as painless. With the catch-up phase past, further growth is likely to be more dependent on improved government performance, specifically increating a favorable investment environment, better administering natural resources, and improving the prioritization and efficiency o f expenditures. 166 See World Bank (2003~). 167 But notethe conclusion o f Chichava et al. (2001) that the decision-making, informationprocessingand analysis functions of the highestlevels of governmentneedimprovement. 108 CHAPTER6: GROWTH AND POVERTYOUTLOOK 370. Mozambiquehas made good progress on the institutional side insome areas. Two examples of growth-enhancinginstitutional change are givenhere: urban water supply'68and mining. Urban water coverage -as measuredby the number of connections - declined from 29 percent in 1992 to 25 percent in2002169owing to the increase o f the urbanpopulation and stagnation inthe number o fhousehold connection due to the delay inthe implementationofthe delegated management policy for urban water supplies. The quality o f service was poor, with an average o f only 11hours o f service per day in2002. Only 19 percent o f standpipes were functional. Cholera cases were common. The government took the decision in 1998 to place urban water provision on a sustainable basis by supporting delegated management of urban systems and engaging incompetitive biddingwith the private sector for a lease contract for Maputo and management contracts for four further major cities -Beira, Nampula, Quelimane, and Pemba. The negotiation processproved to be difficult andwas exacerbated by the devastating effects o f the floods o f 2000. Nevertheless the government stuck by its decision and finally the private sector contractor commenced operations in2002. Within two years it was evident that the right decision hadbeenmade: the number o f people served went up from 33 to 38 percent, despite the 4 percent increase inurban households, and continuity o f supplyhas improved from 11hours to 15 hours on average. Some smaller cities have 20 hours continuous supply which may be interrupted bypower shortages, but water i s available. Furthermore, due to management stability, improved commercial results, and clear progress brought by the private sector, the urbanwater sector has attracted significant additional investments-US$ 30 million for four more cities to be served through management contracts and another USS40 million for investments inMaputo. The lesson leamed i s that even though the institutional change was long and difficult, it was the precondition for long-term progress and has laid the foundation for further improvement inservices and investments leading to economic growth ingeneral. 371. The other example of favorable institutional change i s the Mining Cadastre (see Chapter 5, p. 95). This was created in2003 inorder to establish a non-discretionary and transparent system to grant, manage and cancel miningpermits. On the basis o f the principle "first come, first served", the Cadastre grants mininglicenses based on simple conditions and the payment o f a fixed mining administration fee. As a result the number of new applications for miningtitles increased sharply, from 110 in2001, to 140in2003 and 180 inthe first seven months o f 2004. Also, the average time taken to obtain a license fell from 120 days in 1999 to 50 days in2003. Obviously a Cadastre like this i s not established ina vacuum. There were several other legal, policy and institutional changes that had to precede it. The Government adopted a policy in 1998 whose aim was to establish an attractive socio-economic environment favorable to the development o f the national private sector, and to attract foreign investment. A new MiningCode was introduced in2002 which established an adequate legal framework to attract private investment. State participation inmininghas beenlimited, byprivatizing state-owned enterprises. The capacity o f the officials o f the Ministryo f Mineral Resources and Energyhas beenbuilt up. Artisanal mining has beenencouraged - some 20,000 to 16* See moredetailed discussiono f urbanwater in the Public ExpenditureReview,World Bank (2003b), and a backgroundpaper for the latter by Finney and Kleemeier(2003). 16'See World Bank (2003b), p. 90 for sources and definitions o f coverage. See also paragraph270 (p. 80). 109 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM 50,000 small scale and artisanal miners are estimated to be active inthe country. There i s still much to be done, particularly inreform o f the fiscal arrangements for mining,but the groundwork has been laid for long-term growth inthe sector. 372. These far-reaching institutional reforms permit some hope that similar growth- enhancing reforms might follow inother areas. The shake-up that has occurred inwater mightbe extendedto similar services such as electricity. The reorganization o fallocation procedures inminingmight be repeated, mutatis mutandis, inthe administration o f land inthe large municipalities; landadministration inrural areas; andthe allocation of fisheries licenses. This suggests that with persistent attention to key reforms, Mozambique's potential growth rate could be better than the 5 percent suggested by the educated guesses above, and might rise to 6-8 percent as has been achieved inthe past. B. A BRIEFEXAMINATIONOF THE GROWTHPROSPECTSINKEY SECTORS 373. The major growth sectors in the next decade are expected to be: natural resource based industries (particularly mining), mega-projects and tourism. Miningwas discussed inChapter 5 (page 95) andmega-projects were coveredinChapter 1(page 23). The prospects for tourism are discussed below, as are those o f two other key sectors - agriculture and manufacturing. Agriculture 374. Mozambique's agricultural potential i s scarcely tapped. Only 15 percent o f the country's arable land i s under use currently. Mozambique has 3,300,000 hectares o f land that can be irrigated, of which only 14%i s actually beingproperly exploited. There are tremendous opportunities ininternational agricultural trade inboth cash crops (cotton, tobacco, cashew) and food crops (maize trade with Malawi, rice trade inthe subcontinent, horticulture to EU, to name but a few). Given low domestic purchasing power, substantial growth o f smallholder incomes will require bigimprovements inthe ability to export. Among the key requirements i s financing for commercial ventures and improvements in infrastructure. 375. Yields on field crops are still well below the regional average. Figure 7 shows maize yields o f Mozambique and its neighbors. Despite the remarkable post-conflict catch-up, when yields rose from a nadir o f under 200 k g h a in 1992 to over 900 in 1996, there i s a long distance to go: other countries inthe region - Zambia, Malawi, Tanzania and Kenya -have average yields of 1,450 k g h a due to their superior seed types and greater usage o f modern inputs. Yield growth o f some 60 percent is feasible provided the institutional arrangements can be put inplace - improved extension advice, a trader network for inputs, output markets, and better quality and grading standards, with private sector participation. 110 CHAPTER6: GROWTH AND POVERTY OUTLOOK 2,500 2,000 5 fD T 1,500 E F 1,000 500 ~ ~ ~ S ~ ~ R ~ ~ Figure7. Maize yields, Mozambiqueand its neighbors, 1961-2003170 376. Similarly, large yield improvements couldbe secured inrice, cassava, pigeon peas, groundnuts, and beans. There exists a large regional export market inrice. There i s scope for increasinghorticulture production and export, particularly flowers, baby vegetables, chilies, and tropical h i t s , which couldbe achieved by entrepreneurs working with smallholders inoutgrower arrangements. Contract farming incashew, paprika and other high-value crops could also be pursuedusingthe farmer associationapproach. The outlook for cotton i s less attractive due to low intemational prices and Mozambique's historically extraordinarily low yields (300 kg/ha), but the route o f encouraging farmers' associations has already shown some promise and should be promoted. Considerable scope exists for expansion o f tobacco, provided Mozambique obtains a highly professional institutional capacity to deal with conflicts, achieve quality gains, and ensure the development o f an associative movement which empowers smallholders. " O Source: FAOSTAT(2004). 111 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM BOX 2. SUMMARY OF RURAL DEVELOPMENTSTRATEGY This paper (World Bank, 2005d) finds that growth inthe agricultural economy was primarily due to increases in the area brought under cultivation and (at least up to the late 1990s) increases inthe agricultural labor force as war-displaced individuals returnedto their farms. It stresses the importance o f the smallholder sector and points to tremendous opportunities that exist for expanding trade, particularly inhorticultural exports. It notes that there are serious difficulties in seeking to understand the impact o f government spending, and by implication all donor funding, on the agricultural sector because much spending is off-budget and the budget execution data are insufficiently disaggregated. The mainpolicy advice i s that smallholder productivity should be enhanced, infrastructure access (particularly roads) must be expanded and the costs o f transport reduced, and the Government should invest selectively inirrigation. 377. Policy. The policy environment inMozambican agriculture is much less interventionist than inseveral neighboring countries, Subsidies are almost completely absent. There are a few interventions inthe form o f monopsony marketing areas for cotton, an 18 percent export tax o n raw cashews, a small agricultural credit program (some 60,000 beneficiaries), and an extension service that reaches a minority o f smallholders. The non-interventionist environment i s perhaps good to encourage private initiative, but on the other hand the private sector has not stepped into provide fertilizers, agricultural machinery, seed and so on. This may be due to a demand-side failure, associated with a lack o f mechanisms for rural credit, storage facilities and infrastructure. Whatever the cause, this implies that the risks for smallholders are large. Removing these constraints will be a major task for policy-makers inthe future. 112 CHAPTER GROWTH 6: AND POVERTYOUTLOOK Box3. SUMMARY OF CONTRACT FARMING AND SUPPLY CHAIN FINANCING This paper (World Bank, 20050 on contract farming (or "outgrower arrangements) finds that 400,000 smallholder farmers benefitfrom technical advice and in-kindcredit offered by large-scale firms incotton, tobacco, sugar and oilseeds. A key issue relates to the monopsony concessions -typically an exclusive right to purchase the outputs within a specified distance o f the processing plant. Arrangements o f this kind can lead to abuses and it i s notable that one evaluationo f contract farming inMozambique, Kenya and Zambia found that the share of international prices of seed cotton offered by Mozambique were the lowest. But inthe short to medium term the monopsonies are needed inorder to protect the investment of the processors. The recommendationi s that the monopsonies be subjected to further regulation andthat phased out over a periodo f 10years. Another way o f limitingpossible abuse by monopsony holders i s to hrther encourage rural farmers' associations. Some models innorthernMozambique have shown promise. Ruralbanking i s underdevelopedand the total active loanportfolio i s only $5 million. The avenue o f founding an agricultural development bank i s best avoided because, servicing a restricted and specialized clientele, the viability o f the bank would be compromisedby high fixed costs and over-exposure to risks. A system o f public warehouses could be set up to store agricultural goods, probably cash crops; these deposits could be usedby the owners as collateral for loans. 378. Without close attention to policy there i s the likelihood o f stagnationo f incomes and of a slowdown inprogress towards poverty reduction. There has been some technical improvement, but muchremains to be done by way o f dissemination o f improved varieties and improved mechanical techniques. The low yields could be raised, provided water and road infrastructure improves and as input and output markets grow and skills improve. 379. Furtherroad-building will assist inincreasingmarket participation; it is anticipated that the "Roads IV" program, due to start in 2006, will focus on the tertiary roads needed for agricultural marketing. There are serious problems with input markets which inmany districts are almost altogether absent; though it i s likely that this is due to a lack o f effective demand, and that once the demand i s present, entrepreneurs would come forward to market inputs as they do consumer goods. One o f the constraints on input demand i s the lack o f access to financial markets, which is very limited other than inoutgrower schemes such as cotton; at the very least arrangements needto bemadeto provide ruralpeople with means o f saving. 380. Walker et al. (2004) have shown that the lack o f differentiation o f farming structures i s a constraint on growth and poverty reduction. Simulations show that further differentiation o f the sector resultinginmore medium-sized farms could lead to substantial reductions inpoverty. Relatedto this i s the policy with respect to the large- scale sector in general. It was argued inChapter 5 (p.72) that, while land access i s not a constraint for smallholder farmers, and while the land law offers them reasonable protection, land access for the nascent large-scale subsector i s a problem on account o f the difficulties ofconducting landtransactions. At negligiblylow cost, the large-scale subsector could be encouraged, and this couldhave positive benefits for poverty 113 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM reduction in terms o f the employment created inrural areas, particularly inthe case o f outgrower arrangements. BOX 4. SUMMARY OF IMPACTS OFEXTENSIONSERVICES This paper (World Bank, 2005e) is based on a survey o f 500 farmers executed specifically to examine extension inMozambique. The key finding from a series o f econometric tests i s that extension raises foodgrain output. Inaddition, tests usingdata from the household survey (IAF) showed that extension services are associated with increased household consumption. (Collectively these tests outweigh the negative findings o f Walker et al. (2004) which used the agricultural TIA surveys to show that the reception o f agricultural extension advice had no impact on incomes.) Extension works through the introduction o f new varieties, commercialization andthe promotion o f natural pesticides. The key constraint i s the limited coverage o f the system. The govemment extension service i s not large (700-odd staff) and i s present inbut 52 o f the country's 128 districts, while NGOsprovide extension in42. Only 6 percent o f the rural population lives ina village with an extensionoffice. At least 50 percent o f the rural population has no access to extension services. Furthermore, coverage i s lower inpoorer provinces. Coverage should be extended, and the remit o f extension officers extendedto the dissemination o f information about prices. 381, Finally, there are some sector-specific problems where further policy attention i s needed. The sugar industry, with foreign investment and highprotective barriers, has succeeded invastly expanding output by rehabilitating the plantations o f colonial times, and has created some 6,000 permanent and 19,000 seasonal jobs (Streng, 2004). But despite generous protection, some o f the new plantations are financially troubled. When the world sugar market i s liberalized-probably in less than five years hence - these firms will be unviable. The authorities need to phase in a steady reduction o f the tariff so as to facilitate adaptation. Farmgate prices o f cotton are amongst the lowest in Southern Africa due to lobbying by some o f thejoint venture companies. For equity and growth reasons, the authorities need to introduce countervailing institutional arrangements. Manufacturing 382. Inthe future, thefood andbeverages subsector is likely to grow, drivenby strong internal demand, and backed up by the promise o f vigorous growth inthe tourism and restaurant trades. 383. The woodprocessing andfurniture sub-sector i s also likely to experience healthy rates o f growth, due to the incentives the government i s making available for local processing as opposed to export o f logs. As was noted inChapter 5 (p. 90), with the aim o f increasingvalue added inthe sector, a ban on the export o f first-class round wood was imposed in2003, and a license fee reduction o f 40 percent was granted on logs processed locally. As remarked, better combinations o f incentives could be developed thanthe bluntinstrument of a ban. Irrespective o fthe incentives selected, it is clear that wood processing will increase inthe next decade. 114 CHAPTER 6: GROWTH AND POVERTY OUTLOOK 384. The textile and garments sub-sector. Most o f the developed countries -with the exception o f natural resource-rich countries such as Australia -have gone through a cycle of expanding labor-intensive cheap exports (textiles and garments), followed by heavier manufactures (metal products, etc.) before reachinghighincome levels. This i s also the pattern followed by Mauritius, where 80,000 workers are employed ingarment manufacture, and there are indications that Lesotho, Kenya and Swaziland are following suit (Cockcroft, 2004). 385. Mozambique ought to have certain advantages to enable it to pursue the same path- cheap labor, closeness to the South Afiican market, and trade access under the AGOA and EBA agreements. But textile and garment output declined between 1999 and 2003 (Table 8, p. 20). Inthe late 1990s Mauritian investors contemplated investing heavily inMozambique but finally decided against because the business environment was too unfavorable. Among the key disadvantages, specifically for the textile and garments sub-sector, were the low level o f training o f the work force, the difficulties o f hiring expatriate managers, the highcost o f retrenchment, and slow import and export procedure^.'^^ Dueto the lack o f improvement o f the business climate inthe 1990s and up to 2004, Mozambiquemisseda superb opportunity during which time the AGOA and EBA agreements permittedtariff-free exports. For since January 2005, with the termination o f the Multi-Fiber Agreement, there has been near-free global trade, and with low-cost exports by mainland China it will bemore difficult for Mozambique to compete. There are some remaining tariff preferences which Mozambican exporters can benefit from, e.g.the provision inAGOA than enables transnational companies which locate in Mozambique to use third-country fabric, until 2007.'72 386. There are three stages inthe sub-sector's activity: lint, yarn and garments. At the first stage, currently about 40,000 tons o f lint are produced annually. (The historic peak was 120,000 tons annually.) At the second stage, fabric and yam manufacturing has ceased altogether but there are many mills available, scattered throughout the country, which could be rehabilitated, provided the legal, financial and ownership wrangles could be s01ved.I~~ At the third, one medium sized factory (about 500 employees) accounts for the majorityofgarments exported (Minor 2004), and anew venture funded by the Aga Khan Development, i s settingup a cut and sew operation. Some observers believe that if managed correctly, the sector could employ 600,000 workers (Cockcroft, 2004). 387. Several key steps are neededto rehabilitate the industry: (a) A precondition is that the business environment has to improve dramatically - import and exports arrangements, labor regulation, redtape. Further detail on these has been giveninChapter 4 (p. 47). (b) Efforts could be made to attract several medium-to large-scale foreign apparel firms to invest inMozambique. 17' See Harding and Calvo, 2003, section C. 172 DTIS, 2004, chapter 5. 173 Among the problems are the social claims by workers for compensationwhen the companies ceased production. The old labor law contributed appreciably to strangling the sector inthe early 1990s (Cockcroft, 2004) by requiring single shift work only. 115 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM (c) The government could support the improvement o f value addition at the first and second stages (raw cotton production and spinning). Tourism 388. Tourism i s one o f the fastest growing economic sectors worldwide. Global tourism arrivals reached 692 million in 2001, with a 5.5 percent annual increase forecasted for the next decade. Southern Africa accounted for approximately 1.6 percent o f the total in2001 (10.7 million arrivals), with a projected annual increase o f 7.8 percent over the next few years. Mozambican tourism i s growing swiftly, with annual increases o f over 10 percent recently, for a total o f about 400,000 arrivals annually. Capacity is about 14,000 beds. Its contribution to GDP i s around 1.2 percent. The biggest growth area i s nature-based coastal tourism, inwhich Mozambique has a comparative advantage, due to its natural endowments. The country's extensive coastline (2,700 km) i s unique in the East African marine region interms o f the quality, diversity and species richness o f its habitats. 389. The country has a well-articulated tourism policy and a strategic plan. Recognizing the importance o f the sector for the national economy andpoverty alleviation, the GOM created a separate Ministryfor tourism in2001. Inaddition to its general economic benefits, tourism has a role to play inpoverty reductionbecause (a) the services provided are labor intensive - restaurants, hotels, transport etc. and (b) some o f the areas o f greatest potential are located inareas o f low agricultural potential and with limited opportunities for other types o f income generation. 390. The global tourism industryhas become increasingly competitive as countries have taken measures to attract development. Mozambique tourism faces constraints o f infrastructure and access to highpotential areas, the lack o f institutional capacity to plan and manage tourism development (especially zoning and land use), a poor business climate and a low level o f education and skills (which leads to low productivity and makes the destination price-uncompetitive). The average tourism development entails a relatively small investment o f about US$ 1million. With better planning this could be increased greatly and the resources be put to better use, while attracting more revenue for the exchequer. A comprehensive approach i s needed to address these constraints. In addition, it will be necessary to involve local communities and other stakeholders inthe management of these areas and proceedwith integrated management planning, and specifically to promote o f tourism development that contributes to conservation and sustainable use o f biodiversity. 391. An advantage is that the actual space required for abuilt environment for tourism i s relatively small. It i s proposed that six main areas receive especial focus, starting with Maputo/Inhaca and Vilankulo/Bazaruto Marine Park. 392. Under a "business as usual" scenario, tourism will likely increase its share o f GDP from 1.2 percent to 1.8 percent inthe next decade, and involve employment for a total o f 145,000 people (directly and indirectly). With further reforms, as well as pursuit o f already planned human capital and infrastructure investments, plus some selected public investments, this contribution could rise to 2.8 percent or even 4 percent, and provide as many as 325,000 jobs. It should be noted that the reforms inthe transport sector are already starting to bear fruit: the progressive air transport policy which opened 116 CHAPTER6: GROWTH AND POVERTY OUTLOOK Mozambique's skies enabled the first chartered flight with 200 tourists to arrive from Portugal inJanuary 2005.'74 393. The reforms envisaged are: > Land. The difficulties with the use of land are well-known and have been covered inthe chapter on naturalresource management (p.72). Apart from the reforms called for generally inthe area of landand governance, an additional solution inthe tourism area i s to define special tourism development areas to ensure that sites go to the best developers, through a transparent auction process. > Skills. The skills gap has beencited above. Inthe longterm the educational and skills deficiencies should be addressed on a countrywide basis. Inthe short and medium term, special tourism development areas could help by permitting specific labor training programs, and contracted-in foreign management. > Business environment. The key items inthe difficult business environment are also well known. Pursuing the agenda proposedinChapter 4 on business efficiency (see Table 28, p. 68) will help to address many o f the problems faced by the tourism sector. Among these shouldbe stressed: (a) reducingthe costs and time involvedinbusiness startup; (b) easing the hiringofhigh-level foreign skilled workers, (c) reducing the cost and time involved inretrenchment. C. PROSPECTSFOR FUTUREPOVERTY REDUCTION 394. One o fthe key MillenniumDevelopment Goals i s reducing the poverty headcount by at least halfby 2015. This means halvingthe proportion o fpeople livinginextreme poverty between the dates o f 1990 and 2015, viz. reducing the poverty rate from 69 percent to 34.5 by 2015. To examine whether Mozambique will attain this goal, a simple simulation was prepared, relating growth in GDP to poverty reduction basedon past trends (Table 40). Householdswere dividedinto two groups according to sector of employment o f the head of the household: agriculture and non-agriculture, Consumption o f the household per capita i s assumedto grow at the same rate as GDP inthe sector o f employment o f the head o f household. Growth i s assumed to be distribution neutral. This implies a constant elasticity o fpoverty to growth, and thus the results shouldbe considered upper bound ones, with the likely outcome somewhat lower. 395. It i s assumedthat the non-agricultural sector (68 percent o f GDP and 29 percent o f the population in2002/3) grows at 5.7 percent per annum from 2007 to 2015, and that the agricultural sector grows at 3.2 percent per annum after 2007. A key variable will be how fast households are able to move from the slower growing agricultural sector to the faster growing non-agricultural sectors. Three scenarios are presented: > Scenario 1 uses a 0.8 percent annual population growthrate in agriculture households, which i s very close to the trend inthe IAF surveys, i.e. 0.83 percent, and a 3 percent population growth rate innon-agriculture. These growth rates implythat peoplemigrate rapidly from agriculture to non-agricultural employment. '74 Aghcia Lusa - ServiGo Economia, January 7,2005. 117 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM > Scenario 2 i s calculatedwith the same population growth inboth sectors and the same 2015 population as scenario 1(to obtain this number, population growth i s assumed to be 1.75 percent until 2007 and 1.33 percent thereafter). > Scenario 3 i s calculatedwith the same population growth rates as inscenario 2 but using GDP growth rates which are 1percent lower ineach year inboth sectors. Table 40. Poverty rates, projections for agriculture and non-agriculture Base Scheme 1: sector Scheme 2: zero Scheme 3: zero migra- year migration migration tiodpessimistic growth 2002 2007 2015 2007 2015 2007 2015 Agriculture 58 40 29 45 40 48 48 Non-agriculture 44 40 37 33 20 36 26 All 54 40 31 42 34 45 42 Source: Fox et al. (2005), table 25. 396. Ifgrowthcanbe sustained and inequality wouldremain at more or less the same level, the poverty rate will likely fall to 40 percent by 2007. Inaddition - and subject to the caveat that these estimates are at best rudimentary - it may be possible for Mozambique to reach the Millennium Development Goal for poverty reduction by the year 2015, by reducing its poverty headcount from 54 percent in2002/3 to 31 percent in 2015. But it i s unlikely that inequality will hold steady. Milanovic (1999) examined inequality changes between 1988 and 1993 usinghousehold data sets from 68 poor and rich countries, excluding the ex-Soviet block transition economies. Ofthese, 34 showed increases ininequality and 30 decreases. The average increase was higher (5.3 Gini points) than the average decrease (2.6 points). With this inmind, the projected poverty rate of 31percent by 2015 probably overstates the extent o f poverty reduction. D. KEY RECOMMENDATIONS 397. The full set o f recommendations are given in the substantive chapters 2 to 5 on growth, poverty, private sector and natural resources. Here the key recommendations are brought together so as to demonstrate that they form a credible and coherent whole. 398. The first and most important set o f recommendationshas to do with macroeconomic management, entailing both macroeconomic stability and public finance management. Without low inflation and fiscal control, none o f the structural reforms will succeed inmaintaining strong growth rates andreducing poverty. Given the Government's plans to expand poverty-related spending, the only way that crippling deficits can be avoided i s through raising revenues by persisting with tax reform, broadening the tax base and creating an independenttax agency. 399. Publicjnance management needs dramatic upgrading because Mozambique i s among the poorer performers inAfrica and without substantial improvement, donor largesse which has supported the country's growth will bejeopardized. The SISTAFE should be implementedthoroughly, "off-budgets" brought on budget, budget execution figures reported meaningfully, and civil service pay scales decompressed. 118 CHAPTER 6: GROWTH AND POVERTY OUTLOOK 400. The most important driver of the reductioninpoverty was agricultural growth, as was abundantly shown inChapters 2 and 3. But it was also observed that growth in agriculture has started to plateau out after the post-conflict "bounce-back". Furthermore, there has beenonly modest technological improvement inthe form o f improved seed, use o f fertilizer and machinery. Hence inorder to maintain the pace o f agricultural growth it will be necessary to move to productivity enhancing methods. These will include improved farmer-relevant research, broader dissemination o f drought-resistant varieties, the encouragement o f outgrower arrangements, the encouragement o f farmer associations, and the exploration of new models for access to financial markets by smallholders. 401, While the bulk o f the effort should be inthe smallholder sector - simply because this is where the vast majority ofoutput is produced currently -opportunities also exist forfacilitating investment by medium- and large-scalefarming. Such opportunities should not be neglectedbecause they are low-cost, inthat medium- and large-scale farming operations bringintheir own skills, technology and capital, as witness the tobacco farmers that came from Zimbabwe and set up outgrower arrangements. Attracting medium- and large-scale farming enterprises entails mainly appropriate policy by the Government. Among the measuresproposedinChapter 5 (p. 72) is to increase security o f tenure for medium-and large-scale farming operations by facilitating transactions o f improvements andbuildingsso that sale o f the latter i s automatically accompanied by the land concession on which they are located. 402. Inthe area o fprivate sector development (Chapter 4), this Memorandum did not attempt to break new ground, but sought only to establish a prioritizationfrom among the many measures inthe long pendingagenda. These were dividedinto "quick wins" and "lifting bindingconstraints". 403, Among the "quick wins" was that registering a business should be greatly simplified. This measure is practically costless for the government administration. Among the potentialbenefits-potential becausethe data are not available to predict their magnitude - i s that small firms in urban areas would find it easier and cheaper to register. Their incentives to register are (i) avoid the horse-tradingconsequent upon surprise to inspections, (ii) avail themselves of incentives such as training subsidiesprovided by to the state and (iii) be able to attract better-skilled employees who are lookingfor to inclusion inthe social security apparatus, and (iv) to be able to become supplierso f government and larger firms which are concerned with legality for tax reasons. The probable social benefits from this are that (i) tax take would increase somewhat, (ii) the more firms would be able to benefit from training subsidies, resultinginincreased human capital, and presumably enhanced firm growth, and (iii) corruption inthe public less sector. 404. Inthe category o f "lifting bindingconstraints" i s that electricity supply needs to be more reliable, and access to the network should be increased at a faster rate. This should be done by upgrading the financial and operational performance o f Electricidade de Moqambique, through increased investment,and through institutional and legal reforms to create an enabling environment conducive to the sector's development, with adequate regulatory capacity. 119 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM 405. Inthe area o f naturalresources, the most important intervention, by far, i s to increase safe and sustainable water accessfor rural people. The Government with the backing o f the donor community has had considerable success inpromoting educational services and health services for rural people, but ruralwater access i s the laggard, with coverage of only 27 percent of households. The burdenfalls particularly on women who are obliged to do the fetching and carrying. Increased effort with ruralwater supplywill be required, and there will be a needfor fundingreallocations from sectors o flesser priority. 406. Finally, it is essential that the Government persist with its strategy for combating HIV/AIDS. Mozambique i s one o fAfrica's hardest hit countries, with a prevalence rate o f 16.2percent. The cost to the economy i s o f the order of 1percentage point of GDP per capita each year. Ways needto be found to improve the humancapacity o fthe administration o f the HIV/AIDS programs so that the vast funds which are available are duly spent. 120 REFERENCES REFERENCES Abrams, Leonard, 2005. The role of water in the Mozambique economy: identibing vulnerability and constraints to growth. World Bank, manuscript. April 27. Achcar, Asmara, 2004. Poverty reduction in rural Mozambique: Evidencefrom physical measures of output & non-income measures of welfare: 1996-2002. Mimeo, Partnership (BPRF') . World Bank, February 20. Supported by the Belgian Poverty Reduction AgCnor, Pierre-Richard, Nihal Bayraktar and Karim El Aynaoui, 2004. Roads out of poverty? Assessing the links between aid,public investment, growth, andpoverty reduction. First draft, April 7,2004. Akoroda, M.O., 1999. Study of the contribution of cassava and sweetpotato to totalfood availability in Malawi. U SAID sponsored. May. Alasia, Alessandro, Jan Low and Claudio Soregaroli, 2003. Food demand changes in Mozambique: assessing the role of the rice sector. Center for International Development (CID) Paper Series No. 7. May. Alby, Philippe, Jean-Paul Azam and Sandrine RospabC, 2005. Labor institutions, labor- management relations, and social dialogue. Background paper for the World Bank project "Job creation, core labor standards, and poverty reduction" (initial conference November 17,2004). Andersson, Per-&e, 2001. The impact of the megaprojects on the Moqambican economy. Gabinete de Estudos, MinistCrio do Plan0 e Finanqas, Discussion PapersNo. 18. March. Amdt, Channing, 2003. HIV/AIDS, Human Capital, and Economic Growth Prospects for Mozambique. Africa Region Working Paper Series # 48, February 2003. Amdt, Channing, HenningTarp Jensen, Sherman Robinson, and FinnTap, 1999, Marketing margins and agricultural technology in Mozambique. TMD Discussion Paper No. 43, International FoodPolicy Research Institute. Baker, Dean, Andrew Glyn, David Howelland John Schmitt, 2002. Labor market institutions and unemployment: a critical assessment of the cross-country evidence. Center for Economic Policy (CEPA) Analysis Working Paper 2002-17. November 8,2002. Barro, Robert, 1991. Economic growth ina cross-section of countries. Quarterly Journal of Economics 10612 (May 1991): 407-443. Barro, Robert, and J.W. Lee, 1993. International comparisons o f educational attainment. Journal of Monetary Economics 32(3), December: 363-394. Barro, Robert J., and Xavier Sala-i-Martin, 1995. Empirical analysis o f a cross section o f countries. In:Economic Growth (McGraw-Hill). Pages 414-459. 121 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Benfica, Rui, David Tschirley, and Liria Sambo, 2002. The impact of alternative agro- industrial investments onpoverty reduction in rural Mozambique. Research Report 51E. Maputo. MADERDirectorate of Economics, Departmento fPolicyAnalysis. http://~~w.aec.msu.edula~econlfs2lmozambiquelwps5 1e.pdf Benito-Spinetto, Maria Teresa, and Peter G. Moll, 2005. Macroeconomic developments, economic growth and consequencesfor poverty. (Background paper for the Country Economic Memorandum). February 24,2005, Betcherman, Gordon, Amy Luinstra and Makoto Ogawa, 2001. Labor market regulation: international experience in promoting employment and social protection. Social ProtectionDiscussionPaper Series No. 0128. November. HumanDevelopment Network, The World Bank. Betcherman, Gordon, and Tim Whitehead, 2002. Employment regulation: rulesfor hiring and termination. World Bank Employment Policy Primer. December 2002. No. 1. Bias, Calisto, and Cynthia Donovan, 2003. Gaps and opportunitiesfor agricultural sector development in Mozambique. ResearchReport 54E. Maputo. MADERDirectorate o f Economics, Department o fPolicy Analysis. htt~:~l~~~.aec.nisu.edu/a;econ/Es2lmoza1nbique/~~ps54e.pdf Biggs, Tyler, JohnNasir andRayFisman, 1999. Structureandperjormance of manufacturing in Mozambique. RPED (Regional Program on Enterprise Development). August 1999. Bila, Aniceto, 2003. Master's thesis on the role o f the nonfarm sector inrural development, with specific reference to Boane district. Universidade Eduardo Mondlane. Booz Allen Hamilton, 2002. Assessment of Corruption and Red Tape as Barriers to Trade and Investment in Mozambique. Maputo, December. Bosworth, Barry and Susan M.Collins, 2003. The empirics ofgrowth: an update. For PRMEP (World Bank), and The Brookings Institution. Boughton, Douglas, and DavidMather, 2004. Income growth in rural Mozambique 1996-2002: a challenging comparison: Interim results based on TIA 96 and TIA 2002. Prepared for a technical discussion at the World Bank, November 27, 2004. Boughton, D.,D.Tshirley, H.Manule, A.O. Ofigo, and B. Zulu, 2002. Cotton sector policies in sub-Saharan Africa: Lessons behind the numbers in Mozambique and Zambia. FlashNo. 34E. Maputo: Directorate of Economics, Ministryo f Agriculture and Rural Development. Burnside, Craig, and David Dollar, 1997a. Aid, policies, and growth. PolicyResearch Department, Macroeconomic and Growth Division, WP 1777. Burnside, Craig, andDavid Dollar, 1997b. Aid spurs growth- ina sound policy environment. Finance and Development, December 1997. 122 REFERENCES Bumside, Craig, and DavidDollar, 1998. Aid, the incentive regime, andpoverty reduction. Policy Research Working Paper 1937. Development Research Group, Macroeconomics and Growth, June. Castel-Branco, Carlos Nuno, 2002. Economic Linkages Between South Africa and Mozambique. August 2002. A study for the Department for International Development o f the BritishGovernment - South African Office. Castel-Branco, Carlos Nuno, and Nicole Goldin, 2003. Impacts of the Mozal Aluminum Smelter on the Mozambican Economy. September 21, 2003. Chichava, Ana, Arlindo Lopez, Fernando Pacini, and Jorge Praga, 2001. Relat6rio tkcnico sobre oprocess0 de gestcio depoliticas pziblicas em Moyambique. Estudo realizado no hmbito de implementagzo da reforma do sector p6blico. Setembro. EMCO - Empresa de Consultoria Organizacional. Christie, Iain T., 2004. Overview of tourism in Mozambique. Manuscript. MIGA Swiss Program, World Bank AFTPS. September 14,2004. Chua, H.-B., 1993. Regional spillovers and economic growth. Yale Economic Growth Center DiscussionPaper No. 700. September. Cockcroft, John, 2004. The textile and garment sectors in Mozambique. Submitted by Chemonics International and Nathan Associates, to USAIDMaputo. February. OnNathanwebsite. Coe, DavidT., ElhananHelpmanand Alexander Hoffmaister, 1995. North-South R&D spillovers. NBERWorking Paper 5048. Coelho, Isaias, Paulo dos Santos, Ant6nio Teixeira, Jog0 Amaral Tomas, and Ricardo Varsano, 2001. Para uma Reforma do Sistema Tributdrio e sua Administraycio. IMF,Department ofPublic Finance, March2001. Collier, Paul and Anke Hoeffler, 2002. Aid, policy and growth in post-conflict societies. World Bank Policy Research Working Paper 2902. October. Coulter, Jonathan, Tiago Wandschneider, and Fernando Carvalheira, 2004. Study on contractfarming and supply chainfinancing in Mozambique. August. Natural Resources Institute. For World Bank. Draft final. da Silva, Gabriela Rebello, and Lara da Silva Carrilho, 2002. Bridging the Standards Divide: Challengesfor ImprovingAfrica 's International Market Access. Mozambique Case Study. Dalgaard, Carl-Johan, HenrikHansenand FinnTarp, 2004. On the empirics o f foreign aid and growth. Economic Journal 114 (496): 191. June. Downloaded from www.blackwel1-synergy.com/links/doi/lO. 1111/~.1468-0297.2004.00219.x/fu1l. de Barros, Gilberto, 2005. Private sector development contribution to theMozambique Country Economic Memorandum. AFTPS. World Bank. de Sa, Paulo, 2004. Backgroundpaper on mining for the Mozambique Country Economic Memorandum. September. World Bank. Deaton, Angus, 1999. Commodityprices and growth inAfrica. Journal of Economic Perspectives 13 (3), Summer: 23-40. 123 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Deaton, Angus, and Valerie Kozel, 2004. Data and dogma: the great Indianpoverty debate. September. World Bank, processed. Demographic and Health Survey, Mozambique, 1997. Demographic and Health Survey, Mozambique, 2003. Denison, E., 1967. Why growth rates differ: post-war experience in nine countries. Washington: Brookings Institution Press. Devarajan, Shantayanan, DelfinS. Go,Jeffrey D.Lewis, Sherman Robinson, Pekka Sinko, 1994. Policy lessonsfrom a simple open-economy model. World Bank Policy Research Working Paper 1375. Donovan, Cynthia, 1996. Effects of Monetized Food Aid on Local Maize Prices in Mozambique. Ph.D. Dissertation, 1996, Michigan State University. Available at http://www.aecsnsu.edu/aaecodfs2/niozambique/reprints.htm Donovan, Cynthia, 2004. Agricultural Growth and Poverty Reduction in Mozambique. A seminar paper for aWorldBank workshop, January 13,2004. Supportedby the Belgian Poverty Reduction Partnership (BPRP). Downes, Andrew S., Nlandu Mamingi and Rose-Marie Belle Antoine, 2000. Labor market regulation and employment in the Caribbean. Inter-American Development Bank, Latin American ResearchNetwork, Working Paper No. R- 388. April. Drabu, Haseeb A., 2003. Endof.employment generation. Rediff.com. http://www.rediff.coom/monev/2003/aua/07spec1.htm DTIS, viz. 2004. Removing obstacles to economic growth in Mozambique: a diagnostic trade zntegration study. Three volumes. December 2004. Publishedby USAID. Contractor Nathan Associates. Easterly, W. and R. Levine, 1998. Troubles with the neighbors: Africa's problem, Africa's opportunity. Journal of African Economies 7(l), March. Easterly, William, and Sergio Rebelo, 1993. Fiscal policy and economic growth: an empirical investigation. M E R Working PaperNo. 4499. ECA Labor Market Team, 2004. Labor markets inEurope and CentralAsia: towards creating more and betterjobs. Preliminary draft. November. For the Labor Markets Conference, World Bank,November 18-19. Economic Commission for Africa, 2003. Mozambique -the elusive quest for pro-poor growth. In:Economic report on Africa 2003: accelerating thepace of development. http://www.uneca.org/era2003/chap4.pdf Eicher, Carl. 2002. Mozambique: An Analysis of theImplementation of the Extension Master Plan. Staff Paper No. 2002 31.East LansingMichigan: Department o f Agricultural Economics, Michigan StateUniversity.October. Eide, h e , 2004. An economic analysis of natural resources sustainability in Mozambiquejsheries. August 13. Mimeo, World Bank. Backgroundpaper for the Mozambique Country Economic Memorandum. 124 REFERENCES Eifert, Benn, Alan Gelb and Vijaya Ramachandran,2005. Business environment and comparative advantage in Africa: evidencefrom the Investment Climate data. January. World Bank. Fallon, Peter R., andRobert Lucas, 1991. The impact o f changes injob security regulationsinIndia andZimbabwe. WorldBank Economic Review 513: 395-413. Quoted by World Bank (2003e). Fallon, Peter R., and Robert Lucas, 1996. South African labor markets: adjustment and inequalities. World Bank, Southem Africa Department, InformalDiscussion Papers on Aspects o f the Economy o f SouthAfrica, #12. FEWS (Famine Early Waming System)Malawi, 2004. Informal cross-borderfood trade in Southern Africa. Issue 1, July-September 2004. UnitedNations World Food Program. FIAS, 2001. Mozambique: Continuing to remove administrative barriers to investment. Final, June 2001. IFC. Finney,Chris E., 2003. MozambiquePublic Expenditure Review: backgroundpaper on agriculture. May 2003. Supported by the Belgian Poverty Reduction Partnership (BPRP). Finney,Chris E.,and Elizabeth Kleemeier, 2003. Mozambique Public Expenditure Review: Backgroundpaper on the water sector. May 2003, Final version, sent to DNPO (Ministry o f Finance) and DNI(Public Works) on June 16, 2003. Supported by the Belgian Poverty Reduction Partnership (BPRP). Fischer, Stanley, Ratna Sahay, and Carlos A. Vegh, 1999. Modemhyper- and high inflations. Journal of Economic Literature 6013 (Sept.) Fok, Michel, 1995. The cotton sub-sector in Mozambique: Institutional diversity, performance andprospectsfor improvements. Republic o f Mozambique, MADER(DEA). May. 200 pages. Forteza, Alvaro and MartinRama, 2002. Labor market "ridigity and the success of " economic reforms across more than one hundred countries. For the research project on The Impact o f Labor Market Policies and Institutions on Economic Performance (World Bank). Fox, Louise, Elena Bardasi and Katleen Van DenBroeck, Evolution of poverty and inequality in Mozambique, 1996`7-2002/3. packground paper on poverty for the Country Economic Memorandum). February 17, 2005. Fox, Louise, Gordon Betcherman, Vandana Chandra, BennEifert and Arvil van Adams, 2004. Realizing thepotential of the labor force in Africa: barriers and opportunities. November 16. A summary paper for the Labor Markets Conference, World Bank, November 18-19. Gill, Indermit,Claudio E.Montenegro andDorteDomeland, eds., 2001. Crafting labor policy: techniques and lessons learnedfrom Latin America. The World Bank, Washington, DC. Quoted by Gill et al. (2002). 125 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Gill, Indermit S., William F.Maloney and Carolina Sanchez-Paramo, 2002. Trade liberalization and labor reform in Latin America and the Caribbean in the 1980s. EnBreve, No. 1, May. World Bank. Grobbelaar, Neuma, 2004. 'Every continent needs an America '; the experience of South Africanfirms doing business in Mozambique. The South African Institute of International Affairs, BusinessinAfrica Research Project. Guasch, J. Luis, 1999. Labor market reform andjob creation: the unfinished agenda in Latin American and Caribbean countries. March. World Bank/IBRD. See full text at: http:i/www.inrrentaco1mect.com!content/~~b/331999/00000001/OOOOOOO1/art000 51 - 01 Gwatkin, Davidson R., SheaRustein, Kiersten Johnson, Rohini Pande, and Adam Wagstaff, 2000. "Socio-economic Differences inHealth, Nutrition, and Population". "/Poverty Thematic Group of The World Bank, Washington D.C. Haacker, M.,2001. Providing health care to HIVpatients in Southern Africa. IMF Policy DiscussionWorking Paper, PDP/O1/3, Washington D.C. Haddad, Lawrence et al., 2003. ReducingChild Malnutrition:How Far Does Income Growth Take Us? WorldBank Economic Review 17(1): 107-131. Hall, Robert E. and Charles I. Jones, 1998. Why do some countries produce so much more output per worker than others? Quarterly Journal of Economics 114(1): 83- 116. Harding, Alan and Cesar Calvo, 2003. Mozambican manufacturing in a regional perspective: selected issues and resultsfrom the 2002 manufacturing enterprise suwey. Centre for the Study o f African Economies, University o f Oxford. April. Heckman, James, and Carmen Pagks, 2000. The employment cost of labor market regulations: lessonsfrom Latin America and the Caribbean. University o f Chicago, mimeo. May. Heltberg,Rasmusand FinnTarp, 2001. Agricultural supply response and poverty in Mozambique. WIDERDiscussion Paper No. 2001/114. Herman, Larry, 2003. Mozambique PER: Background Paper on Roads. January 2003. Howard, J., Jose Jaime Jeje, David Tschirley, Paul Strasberg, Eric W. Crawford, and Michael Weber. 1998. What makes agricultural intensijkation profitable for Mozambican smallholders? An appraisal of the inputs subsector and the 1996/97 DNEWSG 2000Program. ResearchPaper Series 31.Maputo. MADER Directorate o f Economics, Department o f Policy Analysis (MADER/DE/DAJ?), October. http://www.aec.n~su.edu/agecon/fs2/moza1nbique!wps31.pdf Hughes, Gordon, 2005. Rural land issues andpolicies in Mozambique. Background paper for the Mozambique Country Economic Memorandum. Hughes, Gordon, 2005. Water issues andpolicies in Mozambique. Backgroundpaper for the Mozambique Country Economic Memorandum. 126 REFERENCES Huppi, Monika, and MartinRavallion, 1990. Thesectoral structure ofpoverty during an adjustment period: evidencefor Indonesia in the mid-1980s. Working Papers: Agricultural and Rural Development Department, The World Bank. WPS 529. IMF (International Monetary Fund), 2003a. StaffReport for the 2003 Article I V Consultation. Appendix I: Relations with the Fund. IMF(Intemational MonetaryFund), 2003b. Ex-post assessment of Mozambique's performance under Fund-supportedprograms. November 3,2003. INE(Instituto Nacionalde Estatistica), 1997. Censo 1997.Maputo, Moqambique. JNE (Instituto Nacional de Estatistica), 2000. Projecgses anuais dapopulacCo economicamente activapor sex0 e idade, 1997-2010. Serie: Estudos N o 3 - Volume 1,Maputo, Moqambique. INE(Instituto Nacional de Estatistica), 2002. Censo de Empresas (CEMPRE), 2002. Tables and data available on the INEweb site. Institute o f Public Finance (Ministry o f Finance, China), 2003. China's employment situation and forecast for the next few years. China & Vorld Economy 4 (2003). Instituto Nacional do Aqhcar, 2003. Balango do PES 2003 no sector do Agzicar. MinistBrio da Agriculture e Desenvolvimento Rural, Maputo. International Horticultural Consultants, February 2002. Prospects of the European marketfor SSA exporters of high valuefresh products. V.E.K.Adviesgroep. February. For World Bank. http://www.tcb-proiect,com/upload files/ Intemational Monetary Fund, 2005, Economic growth andpoverty reduction in Mozambique. Selected Issue Paper prepared by Alvaro Manoel, produced for the Article IV Consultations, May 2005. Islam, Rizwanul, 2004. The nexus of employment growth, employment and poverty reduction: an empirical analysis, Issues inEmployment and Poverty Discussion Paper #14. International Labour Office. January. James, R. C. and C. Amdt, 2005. "Has economic growth inMozambique beenpro- poor?" Maputo, manuscript. Kirkpatrick, Colin, and David Parker, 2004. Infrastructure regulation: modelsfor developingAsia, ADB Institute ResearchPaper Series, No. 60. November. http:/lwww.adbi,oralfiles/2005.O 1.14.rp6O.infrastructure.renulation.asia.pdf Klasen, Stephan, and Silke Woltemann, 2004. The impact of demographic dynamics on economic development, poverty and inequality in Mozambique. Mimeo, World Bank, September 27. Backgroundpaper for the Mozambique Country Economic Memorandum. Supportedby the BelgianPoverty Reduction Partnership (BPRP). Knack, S. and P. Keefer, 1995. Institutions and economic performance: cross-country tests usingalternative institutional measures. Economics and Politics 7(3), November: 207-227. Kostemans, Kees, 2000. Memo for preparation o f Poverty Reduction Strategy Paper: Mozambique. July 2000. Health, Nutrition and Poverty Sector. 127 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM Kugler,Adriana, 2004. The effect ofjob security regulations on labor marketflexibility: evidencefrom the Colombian labor market reform. NBER Working Paper No. 10215. January. Layard, Richard, StephenNickel1and Richard Jackman, 1994. The unemployment crisis. Oxford: Oxford University Press. Levine, Ross and David Renelt, 1992. A sensitivity analysis o f cross-country growth regressions. American Economic Review 8214 (Sept.): 942-963. Lindauer, David L.,1999. Labor market reforms and thepoor. Backgroundpaper for the World Development Report 2000/1, World Bank. Summarizedin Betcherman and Whitehead (2002). Loayza, Norman, Pablo Fajnzylber and CCsar Calderon, 2002. Economic growth in Latin America and the Caribbean: stylizedfacts, explanations, andforecasts. World Bank, June 2002. Macamo, Jose Luis, 1999. Unrecorded cross-border trade between Mozambique and her neighbors. For World Vision International and USAID. SD Publication Series, Office o f Sustainable Development, Bureau for Africa, Technical Paper No. 88. June. McEwen, David, 2004. Transfrontier ConservationAreas (TFCA) and Tourism Development Project (TFCATDP): Study o f economic potential o f tourism in Mozambique. Draft report. December 2004. Ministryo f Tourism (MITUR) and World Bank. Madrid-Aris, Manuel E., 2000. Education 's contribution to economic growth in Cuba. Paper presentedat the annual conference of the Association for the Study o f the Cuban Economy (ASCE), Biltmore Hotel, Miami, August 3,2000. lanic.utexas.edu/lalcb/cubalasce/cubal O1madrid.pdf Margulis, Sergio, 2005. An economic analysis of natural resource sustainability in Mozambique. Background paper for the Mozambique Country Economic Memorandum. April 5,2005. World Bank. Maryam, S., 2002. Analysis of economic growth relationship with employment growth according to the three economic sector inWest Nusa Tenggara (Indonesia). 0yza (Indonesia) 2(1): 33-49. littp://pustaka.bo.gor.net/dtbaru/eko0444.htm Mather, D.,D.Boughton, D.Tschirley and T. Walker, 2005. Analysis of rural income growth in Mozambique 1996-2002 and implicationsfor agricultural development: final results based on TIA 1996 and TIA 2002. Draft presentation. East Lansing: Michigan State University. Processed. Meeuws,Rent, 2004. Mozambique: Trade and transportfacilitation audit. Submittedto the World Bank. September. Milanovic, Branko, 1999. True world income distribution, 1988 and 1993. World Bank, Policy ResearchWorking Paper No. 2244. November. Ministry of Health, Ministry o f Planning and Finance, National Institute o f Statistics and National AIDS Council, 2002. Demographic impact of HIV/AIDS in Mozambique: update 2000. Maputo. 128 REFERENCES Ministry o f Planning and Finance, Eduardo Mondlane University and International Food Policy Research Institute, 1998. Understandingpoverty and well-being in Mozambique: thejrst national assessment (1996-97). Eduardo Mondlane University, InternationalFoodPolicyResearch Institute. December 1998. Ministryo fPlanningand Finance (National Directorateof Planning andBudget, and Economic ResearchBureau), InternationalFoodPolicy ResearchInstitute, and Purdue University, 2004. Understandingpoverty and well-being in Mozambique: the second national assessment (2002-2003). Eduardo Mondlane University, International FoodPolicy Research Institute. March. MinistryofPublic Works andHousing, 2001. Rural water supply project implementation manual. National Directorate o f Water, Rural Water Department. Maputo, December. Minor, Peter J., 2004. A Mozambique textile and garment strategy. Nathan Associates. February 2004. FromNathanAssociates web site http://www.tcb- project.comfupload-files/ . Mole, Paul0 Nicua, 2000. An economic analysis of smallholder cashew development opportunities and linkages tofood security in Mozambique's northern province of Nampula. Ph.D. Dissertation, Michigan State University, East Lansing. http://www.aec.msu.edu/aaecon/fs2/mozan~biqueimoledisse~atio~~.pdf MPF & IFPRI, viz. Ministkrio do Plano e Finanqas, DirecqSloNacional do Plano e Orqamento, and Instituto Intemacional de Pesquisa em Politicas Alimentares, 2003. Pobreza e bem-estar em MoGambique: andise de Questionario de Indicadores Bdsicos de Bem-Estar (QUIBB), 2000-2001. December 2002. Mwangi, Anthony, and Luc De Wulf, 2003, Final report of the survey on customs reform and modernization in Mozambique. March2003. World Bank, draft. Nathan Associates Inc., 2002. Mainstreaming Trade: apoverty reduction strategyfor Mozambique. USAID-funded,part o f the Integrated Framework for trade. Nathan Associates, 2004. Poverty reduction and agricultural trade in sub-Saharan Africa: Recommendedfor USAID interventions. May. Nickell, Stephen, 1997. Unemployment and labor market rigidities: Europe versus North America. Journal of Economic Perspectives 11/3 (Summer): 55-74. Norton, George W., and Jeff Alwang, 1993. Introduction to economics of agricultural development. McGraw Hill. Nunes,Jose C., 2000. Growth, development andpoverty reduction: theMozambique case, Mimeo. EuropeanCommission Delegation, Maputo, draft. O'Connell, StephenA. and Benno J. Ndulu, 2000. Africa's growth experience; afocus on sources ofgrowth. April 16,2000. ww.cid.harvard.edu/cidpeopleibates/research.htm OECD, 1994. OECDjobs study, evidence and explanations. Part Iand 11. Paris: OECD. OECD, 1999. Employment outlook. Paris. 129 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Pal, Parthapratim, 2004. WTO agreement on apculture and its impact on employment inSouthAsia. Paperpresentedat the 5' Annual Global Development Conference, New Delhi, January 2004. Pattillo, Catherine, HelenePoirson and Luca Ricci, 2002a. External debt and growth. IMF WorkingPaper 02169. Pattillo, Catherine, Helene Poirson and Luca Ricci, 2002b. External debt and growth. Finance & Development 39/2 (June). http:llwww.imf.org/extemallpubs/ ft/fandcU2002/06/pattillo.htm Based on Pattillo et al. (2002a). Pavignani, E. (2002). 0 impact0 do SIDA sobre o sector de Saude. Unpublisheddraft. Perumalpillai-Essex, Jeeva, and Asmara Achcar, 2004. Mozambique: key determinants of growth in the agricultural sector 1996-2002. World Bank, mimeo. Supported bythe BelgianPoverty ReductionPartnership (BPRP). Perumalpillai-Essex, Jeeva and Josef Loening, 2005. Mozambique Rural Development Strategy: Chapter 1: Agriculture and ruralpoverty inMozambique: dimension, profiles and trends. April 21,2005. World Bank, draft. Plummer, Janelle, ed., 2003. Better water and sanitationfor the urban poor: good practicefrom sub-Saharan Africa. Water UtilityPartnership for Capacity Building(WLJP)Africa. Abidjan, C6te d'Ivoire andWorldBank, Washington, DC. http://web,mit.edu/urbanup~adin4/wateraiidsanitation/resources/pdf- files/nood practiceslChapter4.pdf Ravallion, Martin and Shaohua Chen, 2003. Measuringpro-poor growth. Mimeo, World Bank. Ravallion, Martin and Gaurav Datt, 1996. How Important to India's Poor i s the Sectoral Composition o f Economic Growth? World Bank Economic Review10: 1-26. January. Reardon, Thomas, Valerie Kelly, Eric Crawford, Thomas Jayne, Kimseyinga Savadogo, and Daniel Clay, 1996. Determinants of farm productivity in Africa: a synthesis of four case studies. East Lansing: Michigan State University. http://~~~.aec.msu.edu/a~econ/fs2/papers/recent.htm#recentidp Regmi,Anita, M.S.Deepak, James L.Seale Jr., and JasonBernstein, 2001. Cross- country analysis of food consumption patterns. May. In:USDA, Changing structure of globalfood consumption and trade1 WRS-01-1. www.ers.usda.govlpublications/wrsOllIwrsOl1d.pdf. Resal (RCseau Europken de SCcuritC Alimentaire) Southern Africa and Food Security Unit, July 2001. Viable optionsfor smallholder crop improvement and diversiJication in Northern Mozambique. Vincent Tickner. From Nathan Associates web site http:!lwww.tcb-proiect.coni/upload filed Riboud, M.,C, Sanchez-Paramo, C. Silva-Jauregui, 2002. Does Eurosclerosis matter? Institutional reform and labor marketperformance in central and eastern European countries in the 1990s. Social ProtectionDiscussion Paper No. 0202, World Bank. 130 REFERENCES Rodrik, Dani, 2003. Growth strategies. Department o f Economics at Johannes Kepler University o f Linz, Working Paper no. 0317, December. Saavedra, Jaime, and Maximo Torero, 2000. Labor market reforms and their impact on formal labor demand andjob market turnover: the case of Peru. Research Network Working Paper No. 394. Inter-American Development Bank, Research Department, Washington, DC. Quotedby World Bank (2003e). http://www.iadb.or~res/~ublicatioiis/p~~~~les/pubR-394.pdf Sachs, Jeffrey D., and Andrew M.Warner, 1997. Sources o f slow growth inAfrican economies. Journal of African Economies 6(3): 335-376. SAL Consultoria e Investimentos, 2002. USAID/Mozambique Country StrategicPlan (CSP): assessmentof Mozambican labor law. Maputo. October. See full text at www.tcb-proi ect.com/ SAL Consultoria e Investimentos, 2003. The legal and administrative framework for labor relations inMozambique. December. Fundedby the United States Department o f Labor. Sarkar, Soumodip, 2000. Mozambique too can compete! I n labor-intensive manufactured exports. Ministry o f Planning and Finance, Gabinete de Estudos, Discussion Paper 14 (November 2000). Seroa da Motta, Ronaldo, 2004. An economic evaluation of forestry regulation in Mozambique. November. Backgroundpaper for the Mozambique Country Economic Memorandum. SIMAteam. 2004. Quente Quente # 447, January 28,2004. Maputo: MADER/ DEST / SIMA. http://www.aec.msu.edu/aaecoilifs2/inozambique/weeklybulletin.litin Solow, Robert, 1957. Technical change and the aggregate production function. Review of Economics and Statistics 39: pp. 312--320. Sousa, Daniel, Josef Loeningand Jeeva Perumalpillai-Essex, 2004. Agricultural data in Mozambique: sources and issues. Mimeo. October 12, 2004. From Back To Office Report - Visit to ruralNampulaNiassa provinces and agricultural data issues. Sowman, Ray, 2001, SADC regionalfreight transport corridors: commodities, tonnages, andfreight per mode. August. Prepared for Rapid Project. A USAID-Rapid Project. 76 pp. Steckel, Richard, 2004. Education and economic growth. Economics.sbs.ohio- state.edu/rsteckel Streng, Peter, 2004. Human capital development: Mozambican sugar industry. Presentation - the World Bank & donor community: Maragra Apicar. June 18, 2004. Sub-sector cotton study, 2001. http://www.tcb-proiect.com/upload files/ (Nathan Associates web site). October. Draft report. 131 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM Tahari, Amor, Dhaneshwar Ghura, Bemardin Akitoby, and EmmanuelBrou Aka, 2004, Sources of Growth in Sub-Saharan Africa. IMF Working Paper No. 041176 (Washington: International Monetary Fund). Tarp, Finn, Channing Arndt, Henning Tarp Jensen, Sherman Robinson, Rasmus Heltberg, 2002. Facing the development challenge in Mozambique: an economy-wide perspective: an economywideperspective. IFPRI ResearchReport 126. Technoserve, 2002. Briefingdocument: the Mozambican h i t industry. October. Technoserve, 2003. Assessing the competitiveness o f the horticultural sector inManica Province. October 17. Interim(Phase I) report for the IFC. Thirlwall, A.P., 1983. Growth & development with special reference to developing economies. London: Macmillan. Tickner, V., FinnRisberg,Marcos Freire, and James E. Cooper, 2001. Viable optionsfor smallholder crop improvement and diversiJication in northern Mozambique. Maputo: RESAL Southern Africa and EuropeanCommission Food Security Unit Mozambique, July. http:llw~~~.delmoz.cec.eu.int/enicooperationlsectorsifood evaluation.htm Toyoshima, Toshihiro and Yutaka Yoshino, 2004. Country Analytical Briefs. UsingUN COMTRADE data. June 2004. Processed, World Bank. Treichel, Volker, 2005. Tanzania's Growth Process and Success in Reducing Poverty. IMFWorking Paper No. 05/35 (Washington: InternationalMonetary Fund). Tschirley, David L.,and Ana Paula Santos, 1999. The Effects of Maize Trade with Malawi on Price Levels in Mozambique: Implicationsfor Trade and Development Policy. Ministry o f Agriculture and Fisheries Research Paper series No. 34. 20 November 1999. http:l/www.aec.msu.ed~iia~econlfs2lmo~ainbique/wps34.pdf Tynan, Nicola, and Bill Kingdom, 2002. A water scorecard. Public policy for the private sector. NoteNumber242, April. The World Bank Group: Private Sector and InfrastructureNetwork. rm.worldbank.org/PapersLinks/Open.aspx?id=l985 UNDP, M0zambique:National Human Development Report 2001. Maputo. UnitedNations, 2001. An investment guide to Mozambique: opportunities and conditions. UNCTAD-International Chamber o f Commerce. New York: June. UnitedNations Population Division. 2004. WorldPopulation Prospects Database. New York: UnitedNations. http://esa.un.orFI/unpp/ USAID,n.d.(about 2000). Thepotential of the tourism sector in Mozambique to impact poverty. From Nathan Associates web site. Verbeek, Jos, 2003. Poland: macroeconomic developments, economic growth, and consequencesfor poverty. October 23, 2003. Draft. Walker, Jane, 2005. Maputo municipal development: Water and sanitation - experience, success and challenges. March 8. Presentation at the UrbanForum, Washington, DC. (Data from FIPAG.) http:/,'www.worldbank.orrr/urbaniurbanforum20~5/prcsentations/walker.pdf 132 REFERENCES Walker, T., D. Tshirley, J. Low, M.Pequenino Tanque, D.Boughton, E.Payongayong, and M.Weber, 2004. Determinants of rural income, poverty, andperceived well- being in Mozambique in 2001-2002. Ministry o f Agriculture and Rural Development: Economics Directorate. Research Report No. 57 E. May. http://~~~.aec.m~~.ed~~a~ec0nlfs2/mozambique/wps57e.pdf Wandschneider, Tiago Sequeira and Julio Garrido-Mirapeix (1999). Cash Cropping in Mozambique: Evolution and Prospects. Technical Paper 2. Maputo: Food Security Unit inMozambique, EuropeanCommission. http://europa.eu.int/comm/europeaid/proiects/resal/them/theme3 .html#6 Whiteside, Martin, 1998. m e n the whole is more than the sum of theparts: the efect of cross-border interactions on livelihood security in southern Malawi and northern Mozambique. December. A report for Oxfam GB. World Bank, 1997. Mozambique: Agricultural Sector Memorandum. Volumes I: Summary (39 pages) and 11:MainReport (180 pages). Africa Region, Agriculture and Environment, World Bank (AFTA1). Report No. 16529 MOZ. April 30. World Bank, 1992. Mozambique -SecondPublic Expenditure Review. December 3 1, 1992. Report No. 11524 -Moz. World Bank, 1997. Memorandum and recommendation of the President: Road Sector Investment Program (ROADSIP). September 4. Report No. P-7177-ZA. World Bank, 2001a. Malawi. Public Expenditures: Issues and Options. Report No. 22440 MAL Grey Cover. September 2001. Africa Region: Macroeconomics 1. World Bank, 2001b. Zambia: Public Expenditure Review: Public expenditure, growth andpoverty: a synthesis. Report No. 22543-ZA. December 2001. World Bank, 2001c. Project Appraisal Document on a Proposed Credit in the amount of SDR 11.6million ($14.9m)for Mozambiquefor a Communication Sector Reform Project. Oct. 30,2001. Report No. 22956-MOZ. World Bank, 2001d. Mozambique Country Economic Memorandum: growth prospects and reform agenda. Report No. 20601-MOZ. February 7,2001. Macroeconomics I, Africa Region. World Bank, 2002. Implementation Completion Report on the Flood Emergency Recovery Project. Report No. MZ-24415. December 19,2002. Task Team Leader Dipac Jaiantilal. World Bank, 2003a. Mozambique: Industrial Performance and Investment Climate. John Nasir, ManjuKedia Shah, Chad Leechor, Vijaya Ramachandran. World Bank, 2003b. Public Expenditure Review, Phase 2: Sectoral Expenditures. World Bank. Report No. 25969. September 22. World Bank, 2003c. Mozambique: Legal andjudicial assessment. Draft. March. World Bank, 2003d. Update on Implementation of Action Plans to Strengthen Capacity of HIPCs to Track Poverty-Reducing Public Spending. March 7, 2003. Board paper o f World Bank and IMF. See 133 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM www1.worldbank.org/publicsector/pe/HIPCUpdate.pdf,www- wbweb.worldbank.or,g/prem/preinps/expetiditure/hipc.htm. World Bank, 2003e. Doing business in 2004. Washington, DC. rru.worldbank.or~Docuinents/DoinBusiness/2004/db2004-chapter-3 .pdf World Bank, 2004a. GEFProject Briej Mozambique: Transfrontier ConservationAreas and Tourism Development Project. Draft. World Bank, 2004b. Project appraisal document on aproposed creditfor a Road Rehabilitation and Maintenance Project. February 13, 2004. ReportNo. 27891, World Bank, 2004c. Doing business in 2005: removing obstacles to growth. Washington, DC. World Bank, 2004d. Health Country Status Report. February23, 2004. World Bank, 2005a. Poverty and Social Impact Analysis: Primary School Enrollment and Retention - the Impact of School Fees. Report No. 29423-MZ. World Bank, 2005b. The role of water in the Mozambican economy - identzhing vulnerability and constraints to growth. May 24, 2005. Memorandum. Task Team Leader: LeonardAbrams. World Bank, 2005c. Growth and competitiveness. June 22,2005. Private Sector Development,AFTPS, Afi-icaRegion. White cover draft. World Bank, 2005d. Rural development strategy. Yellow cover version, June 20. Report no. 32416-MZ. Environment,Ruraland Social Development Department, AFTS1. World Bank, 2005e. Impacts of extension sewices in rural Mozambique. ECON Analysis andthe World Bank. Mimeo. World Bank, 2005f. Study on contractfarming and supply chainfinancing in Mozambique. Draft, January 31,2005. Task team leader Eduardo L.Le20 de Sousa. AFTS1. Yagci, Fahrettin, 2004. Mozambique: salientpointsfor a trade strategy. Internal note. October. Yagci, Fahrettin, 2004. Comments on ShouldMozambiqueJoin SACU?A Review o f Kev Issues (A report preparedby TSG for the Governmentof Mozambique). July 12. World Bank,processed. Yagci, Fahrettin, and EnriqueAldaz-Carroll, 2004. Salientfeatures of tradeperformance in Eastern and SouthernAfrica. Africa RegionWorking Paper Series No.76, October 2004,2004. Younger, Geoffrey Gordon, and The BeefCompany Pty. Ltd., 2003. Mozambique rice project: Business Plan. May. 134 ANNEX1:TABLES AND FIGURES ANNEX 1 TABLES AND FIGURES Annex Table 1. Selected economic and financial indicators175 Actual Prelim Pro- jected Indicators 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Output and prices(annual growth) RealGDP 4.3 7.1 11.1 12.6 7.5 1.9 13.1 8.2 7.8 7.2 7.7 7.4 6.4 6.4 Real GDP per capita 1.7 4.4 8.4 10.4 5.5 -0.3 10.7 6.0 5.8 5.3 5.9 5.7 4.7 4.6 Realconsumptionper capita -19.7 6.7 9.5 4.3 4.2 -2.9 -3.8 6.1 8.0 6.6 7.4 2.4 4.0 3.1 Consumerprice index (ann. a%.) 54.4 44.6 6.4 0.6 2.9 12.7 9.0 16.8 13.4 12.6 8.0 7.3 6.5 5.9 External sector (US% m) Imports(c.i.f.)atcurrentprices 727 783 760 817 1,200 1,163 1,063 1,543 1,741 2,035 2,353 2,419 2,661 2,737 Exports (f.0.b.) at current prices 174 226 230 245 284 364 703 810 1,044 1,504 1,713 1,746 1,778 1,889 Exportvolumeindex(l992=~00) 113 148 152 191 232 277 577 671 780 887 925 1018 1139 1236 Import volume (index) 90.9 100.4 106.6 123.2 182.0 177.6 170.0 239.5 240.1 251.3 271.5 279.7 307.6 314.1 Change in terms oftrade(dec1 - ) 0.3 1.8 7.7 -8.7 -4.1 8.3 -2.9 -3.8 -1.4 13.4 2.2 -7.3 -8.9 -2.8 Real eff. exch. rate (depn. -) -4.4 14.8 9.4 -7.7 1.6 -4.1 -9.4 -6.4 -2.8 24.8 .. Money and credit (trillion Mt) Yet domestic assets 8.4 7.6 8.3 9.0 11.1 12.4 14.6 15.1 17.8 10.1 14.8 19.3 22.5 25.7 Net domestic credit 0.9 0.5 1.0 1.1 3.1 6.2 9.7 10.9 10.8 9.2 14.5 20.6 25.8 28.0 of which: Govemment (1.6) (3.1) (4.4) (5.6) (5.6) (5.1) (4.3) (3.6) (3.5) (4.3) (1.0) 1 5 4.4 4.4 Privatesector 2.5 3.6 5.4 6.7 8.7 11.3 13.9 14.5 14.3 13.5 15.5 19.1 21.4 23.6 money andquasi-money(M2) 4.9 6.0 7.4 8.7 11.8 16.8 22.4 27.2 32.3 34.2 39.1 44.7 50.6 57.6 M 2 growth rate (percentage) 21.1 24.4 17.6 35.1 42.4 33.2 21.5 18.7 5.9 14.5 14.2 13.3 13.6 Public finances (% of GDP) Generalgovt revenue(excl grants) 11.7 10.6 11.6 11.3 12.0 12.9 12.4 12.4 12.9 12.3 13.2 14.0 14.6 15.1 Tax revenue 10.6 9.8 10.6 10.5 11.0 11.8 11.0 11.0 12.0 11.3 11.8 12.5 12.9 13.4 Nontax revenue 15.4 13.5 16.0 15.4 18.3 17.1 16.9 15.8 16.6 15.1 15.4 16.0 16.5 17.0 Expenditure 24.9 20.7 23.9 21.6 24.7 26.6 32.1 30 0 26.5 23.7 25.6 24.2 24.5 24.8 Current expenditure 10.6 9.4 10.7 11.2 12.2 13.1 13.5 13.9 14.4 13.8 13.7 13.6 14.0 14.1 Of which: wages and salaries 2.4 2.2 3.6 4.5 5.8 6.5 6.5 6.4 6.8 6.7 6.9 6.9 6.9 7.0 Capital expenditure 14.4 11.3 13.1 10.4 12.5 13.5 18.6 16.1 12.2 9.9 11.9 10.6 10.5 10.6 Overall deficit before grants (-) -13.3 -10.1 -11.9 -10.5 -13.2 -13.7 -19.9 -17.3 -14.0 -11.7 -12.4 -10.2 -10.0 -9.7 Overall deficit after grants -3.2 -3.1 -2.9 -2.2 -1.0 -5.8 -6.0 -7.2 -4.5 -4.4 -6.0 -4.4 -4.4 -4.2 Primary balanceafter grants -1.5 -1.6 -1.6 -1.2 -0.4 -5.6 -5.4 -5 9 -3.3 -3.4 -5.2 -3.5 -3.3 -3.1 Savings and investment Gross domestic savings 5.0 5.4 8.1 10.8 13.7 11.6 8.0 11.0 10.1 11.8 13.2 14.1 14.2 14 5 Gross dom fixed capital formation 30.5 21.8 20.6 24.2 36.7 33.5 25.9 29.8 25.9 20.1 21.3 21.3 23.8 22.5 Public sector 12.4 10.3 12.1 9.8 11.6 10.4 15.4 12.5 11.7 9.1 10.8 9.8 9.7 9.9 Privatesector 18.1 11.6 8.5 14.5 25.1 23.2 10.5 17.3 14.2 10.9 10.5 11.5 14.1 12.6 Extemal currentaccountbalance -29.1 -22.3 -17.7 -18.9 -28.4 -27.2 -26.1 -23.1 -19.9 -13.8 -13.9 -12.9 -15.0 -13.1 Other indicators Gross official reserves Inmillions ofUS$ 225 383 532 625 669 745 727 825 947 1159 1076 1045 1104 1201 Inmonthsofimports ofg&s 3.O 4.8 6.8 7.1 5.2 5.8 5.1 5.4 6.0 5.1 4.O 3.7 3.5 3.7 Exchangerate (MUUS$) 8,890 11,294 11,544 11,875 12,775 15,227 20,704 23,678 23,782 22,581 Current GDP (in US$billions) 2.3 2.9 3.4 4.0 4.1 3.8 3.7 4.1 4.8 6.1 6.5 7.0 7.5 7.9 NPV extemal debt / exports -1358 -709 549 212 177 110 92 102 84 84 90 --96 99 Source: Live DataBase, inturn from INE, MPF, IMFand World Bank staff missions. 135 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM Annex Table 2. Mozambique's progress towards the MillenniumDevelopment Goals 176 MillenniumDevelopment Goal Will the Comments goal be met? Extreme poverty: halve the proportion of - - Possible Rate fell from 69% in 199617 to 54% in2002i3. people living inextremepoverty between Projecting, may be as low as 31% in2015 - subject to 1990-2015,viz. from 69% to 34.5% the qualifications cited inthe maintext. HIVMIDS:Halt and reverse the spreadof Unknown Prevalence rate rose from 12 percent in2002 to 16.2 HIViAIDS by 2015 percentin2004. Hunger: Halve the proportion of people Unknown who suffer fromhunger by 2015 Access to safe water: Halve the proportion Rural: See maintext (p. 80), and World Bank (2003b). o f peoplewithout safe drinking water by possible Projecting the same rate of increase as that observed 2015 (viz. rural: raise access from 12%to Urban: from 199617 to 200213 would give 58% rural and 56%, urban: from 56% to 78%). possible 8 1% urban coverageby 2015. Note: The base year for all the MDGs is 1990. We assume no change between 1990and 199617. Education:Universal completion of Unlikely Completion raterose from say 15%("less than 20%" primary educationby 2015 accordingto World Bank, 1992, p. 84) in 1990to 36% in2003. Projecting, it will be 51% in2015. Unlikely to be met despite Education For All due to constiaints on recurrent expenditure. See World Bank(2003b). Gender equality: Eliminate gender Yes Access: Achieved w.r.t.primaryby 2002. disparity in: . (Completion: likely date unknown; depends on efforts primary educationby 2005 to recruit female teachers.) . secondary educationby 2005 unknown Access: likely date unknown. (Completion: likely date unknown.) all levels of educationby 2015 Unknown Access at tertiary: likely date unknown. (Completion: likely date unknown). Child mortality: reduceunder-5mortality Yes IMR droppedfrom207 per 1,000 in 1993-8to 153 in by 213 between 1990and 2015, viz. from 1998-2003 (World Bank, 2004d, table 3, p.23). 22611000to 7511000 Projecting, the rate in2015 i s 76 (idem, Figure 6). Maternal health: reduce maternal Yes MMR was 408 in2003 (World Bank, 2004d, para. mortality by % between 1990and 2015, lo), not farfromthe MDGtarget. viz. from 980/105live births in 1995 (World Bank, 2004d, para. 50) to 365110' (World Bank, 2004d, para. 10) Malaria:Havehaltedby 2015, andbegun Unlikely No evidencethat the actions takenhitherto have to reverse, the incidenceof malaria reducedincidence Environment: Integratethe principles of Poten- n.a. sustainable developmentinto country tially policies andprograms andreducethe loss of environmentalresources 176 For generalinformation about the MDGs, see see http://www.undp.orgimdg/countryreports.html. 136 ANNEX 1: TABLESAND FIGURES Annex Table 3. Employmentby employer type, 199617 and 2002/3 ( O hof labor force) 199617 200213 Rural Urban All Rural Urban All Governmenb`public enterprise 2.7 26.5 6.0 1.4 11.5 4.2 Private sector 2.2 14.3 3.9 1.8 12.7 4.8 Cooperative sector 0.1 0.4 0.1 0.0 0.2 0.1 Self-employment 46 40 45 48 40 46 Employer 0.2 1.o 0.3 1.4 10 3.6 Family labor 49 18 45 47 26 42 Total 100 100 100 100 100 100 Source: IAF 199617and2002/3, workings for the accompanyingbackgroundpaper by Fox et al. (2005). Note that the 199617definition of "urban" was usedinboth 199617and 200213. This is becausethe urban areas were extended by some 50 percentinthe intervening period, makingthe two definitionsincomparable. Age group: 10-59years. Annex Table4. GDP per capita,and GDP sharesby province/region,2000 GDP per capita, GDP current US$ share, 2000 1996 2000 Niassa 96 105 3 Cab0 Delgado 112 105 5 Nampula 130 137 14 Zambezia 102 85 9 Tete 98 124 5 Manica 121 130 5 Sofala 247 193 9 Inhambane 135 118 5 Gaza 112 101 4 Maputo Province 122 2 14 3 Maputo City 1079 1137 37 Mozambique 185 188 100 Source, UNDP (2001), p. 25, graph 2.8. 137 MOZAMBIQUECOUNTRY ECONOMICMEMORANDUM Annex Table 5. Growth of GDP by province/region, 1997-2000 1997 1998 1999 2000 Average North 19.0 10.5 3.5 0.9 8.5 Niassa 18.7 3.9 11.5 6.9 10.3 Cab0 Delgado 1.5 23.4 3.9 -1.3 6.9 Nampula 25.9 7.6 1.9 0.5 9.0 Center 10.0 7.1 10.2 -3.4 6.0 Zambezia 7.1 -3.4 8.1 0.8 3.3 Tete 19.8 1.1 12.5 16.6 12.5 Manica 21.9 28.1 11.9 -11.1 12.7 Sofala 4.3 10.3 10.2 -10.5 3.6 South 8.3 17.2 7.8 4.8 9.5 Inhambane 8.3 8.4 10.0 -1.3 6.3 Gaza 19.1 14.1 9.5 -17.1 6.4 Maputo Province 45.3 5.6 12.2 26.5 22.4 Maputo City 3.5 20.5 6.8 5.9 9.2 Mozambique 11.1 12.6 7.5 1.6 8.2 $ 0 Maputo 1 5 - province m Maputo city 0 b 2c o - ~ o -5 I 0 Sofala 1 10 100 GDP per capita, 1996 (current Mt 1OA6,logscale) Annex Figure 6. GDP per capita in 1996 and annual growth of GDP per capita 1996-2000, by province'77 177 GDP per capita in 1996: in current Mt. Growth refers to average annual growth inthe period 1996- 2000. The CPI was usedto exclude the effect o f inflation. Data source: U N D P (2001). 138 ANNEX1:TABLES AND FlGURES 1,000,000 10,000,000 GDP per capita, 1996 (current Mt, logscale) Annex Figure7. GDP per capita in 1996 and annualapprovedinvestment 1996-2000 as a percentageof GDP, by pr~vince"~ Annex Table 8. Macroeconomic impactof the mega-projects,1998-2002 1998 1999 2000 2001 2002 Avg98- 02 M i l l i o n s of USD u n l e s s o t h e r w i s e i n d i c a t e d Value added 13 67 80 183 264 122 as % o f GDP 0.3 1.6 2.1 5.3 7.3 3.2 Fiscal accounts 0.0 0.3 1.7 2.0 5.4 1.6 as % of fiscal revenues 0.1 0.3 0.4 1.1 0.4 Overall BOP 13 64 91 202 197 113 as % o f GDP 0.3 1.6 2.4 5.9 5.5 3.O Exports of goods 0.0 76 127 441 437 216 as % of total exports of goods 28 35 63 64 48 Source: See Benito-Suinettoand Moll (2005). GDP per capita in 1996: incurrent Mt,log scale; data source: UNDP (2001). Approved investment excludes the Mozal investment in the case o f Maputo province, and is taken as a percentage o f GDP o f the relevant province; data source: from Table A.4 inWorld Bank (2003a), p. 60, in turn from data provided by the Centro de PromoqBo de Investimentos. 139 MOZAMBIQUE COUNTRY ECONOMIC MEMORANDUM Annex Table 9. Projectedmacroeconomicimpactof the mega-projects (2003- 2010) Avg.03- 2003 2004 2005 2006 2007 2010 10 I n m i l l i o n s o/ US% u n l e s s o t h e r w i s e i n d i c a t e d 1. Including existing projects Value added 352 484 560 653 708 729 618 as % of GDP 8.2 10 10 11 11 8.7 10.0 Fiscal accounts 7.6 13 36 40 42 44 33 as % of fiscal revenues 1.2 1.9 4.3 4.3 4.0 3.1 3.0 Overall BOP 219 161 221 267 320 282 269 as % of GDP 5.1 3.3 4.1 4.5 4.9 3.4 4.0 Exports o f goods 652 963 1106 1258 1338 1367 1176 as % of exports of goods 71 76 74 78 77 76 75 2. Including existing and new projects Value added 352 484 584 670 807 961 761 as % of GDP 8.2 10 11 11 12 11 11 Fiscal accounts 7.6 13 36 40 50 71 48 as % of fiscal revenues 1.2 1.9 4.3 4.3 4.8 4.8 4.5 Overall BOP 219 160 244 285 370 387 328 as % of GDP 5.1 3.3 4.5 4.8 5.6 4.5 4.9 Exports o f goods 652 ,963 1106 1258 1498 1941 1475 as % of exports o f goods 71 76 74 78 79 82 79 Source: See Benito-Spinetto and Moll (2005) Annex Table 10. Distributionof projectizedforeign aid by sector I 1995 1996 1997 1998 1999 2000 2001 2002 P e r c e n t a g e Agriculture 17 11 14 14 16 14 23 16 Fishing 1.3 0.6 1.2 0.8 1.3 0.6 0.4 0.6 Mineral resources 0.8 1.6 0.5 0.3 0.3 0.2 0.4 0.8 Industry and Commerce 2.0 1.3 0.7 1.1 0.4 0.3 0.5 0.3 Public works and housing, olw: 24 16 17 17 14 17 15 18 Water 5.6 3.5 2.8 3.9 3.1 6.3 5.5 4.0 Roads 18 12 14 13 11 11 9.3 14 Other public works 0.8 0.5 0.2 0.1 0.0 0.2 0.2 0.0 Transport 7.1 4.1 3.2 4.4 3.8 2.5 2.1 7.0 Energy 6.2 14.9 11.0 6.3 9.9 4.5 7.4 7.3 Telecommunication 2.1 2.8 1.9 4.0 1.2 2.0 2.3 3.0 Educationand culture 8.4 9.2 12.2 9.3 11.9 9.8 10.1 9.9 Health 11 8.9 9.6 9.3 11 12 17 14 Tourism 4.1 4.9 4.4 4.5 7.1 5.4 2.0 4.3 Environment 0.4 0.6 1.3 3.2 3.1 1.8 1.0 1.5 Public administration 5.6 8.3 8.0 6.0 6.6 5.3 5.8 7.3 Social 0.2 0.4 0.6 0.2 0.6 0.4 0.7 0.3 Other 9.4 16 15 19 14 24 13 9.4 Total 100 100 100 100 100 100 100 100 Sources: Departamento de Coopera@o Intemacional, Ministry of Planning and Finance. 'Aid here includes grants and borrowing. It excludes program assistanceas this cannot be assigned to specific sectors. 140 cr W 0 N r- N N - 0 1 0 3 - m r - m * N m 2 * - " 3 c4 m vi m - m m m e r - - m m d cma rT- t " w d d M M cam b m l ' \o d m m m o - v i m m m 3 r- - d c\ N vi d r- m a r - mmm: vi & - 3 3 M d "rg ' v i d ' ' 3 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM growth incidence curve 0 20 40 60 80 100 Percentiles I Median spline Mean of growth ratesI Annex Figure 14. Growth incidence curve, 199617to 200213 144 ANNEX 1:TABLESAND FIGURES Inequality Subsaharan Africa National level Mozambique (1996/2002) Mauritania (1995/2000) Mali (1994/1999) Ghana (1991/1998) Uganda (1992/2002) Kenya (199411997) CBte d'ivoire (199311998) Cameroon (1996/2001) Madagascar 1993/2001) Burkina Faso I994/1998) Ethiopia 1996/1998) Nigeria 19931996) Zambia 1996/1998) 0 0.1 0.2 0.3 0.4 0.5 0.6 Gini coefficient Annex Figure 15. Inequality in sub-Saharan Africa 145 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Annex Table 16. Decomposingchanges in povertyinducedby changesin mean consumptionand inequality(by geographicaland sectoraldimensions) Total change Change in mean Changein Residual inpoverty consumption inequality National decomposition Total change inpoverty 2002-1996 -15.1 -16.9 1.3 0.5 Regional decomposition Change inpoverty inthe North -10.5 -11.7 3.4 -2.3 Change inpoverty inthe Center -28.2 -28.7 2.2 -1.7 Change inpoverty inthe South 0.7 -2.2 2.9 0.04 Regional+ urban decomposition Change inpoverty inNorth urban -40.5 -33.1 -11.4 4.0 Change inpoverty inNorth rural -5.1 -6.3 4.2 -3.0 Change inpoverty in Center urban -20.3 -25.3 2.3 2.8 Change inpoverty inCenter rural -29.2 -29.0 2.3 -2.5 Change inpoverty inSouth urban 7.1 -1.2 8.6 -0.4 Change inpoverty in Southrural -3.2 -2.8 -0.9 0.5 Urban-rural Change inurbanpoverty -10.1 -11.3 2.0 -0.8 Change inrural poverty -15.8 -15.7 -0.6 0.5 Urban-rural (consistent definition) Change inurbanpoverty -12.1 -13.6 1.4 0.2 Change inruralpoverty -15.9 -18.4 1.4 1.1 Aggregate sectors Change inagriculture poverty -14.4 -13.0 -0.7 -0.7 Change inindustrypoverty -8.9 -19.3 6.9 3.5 Change inservice1 poverty -9.2 -11.2 -0.8 2.7 Change inservice2 poverty -19.9 -22.6 0.8 1.9 Head employment status Head is public employee -24.9 -28.8 -0.1 4.1 Head is private employee -12.5 -20.4 4.2 3.6 Head self-employed -14.6 -15.1 0.3 0.2 Head is employerico-operative 2.8 11.0 -14.2 6.0 Head infamily business -20.9 -16.3 0.5 -5.1 Source: IAF 199617 and 200213! as computedby Fox et al. (2005). Individuals are assignedto the sector of employmentof the householdhead. Ifthe head is not employed they are assigned to the sector of employment of the oldest adult. Ifnobody works (less than 5% of all cases) they are assigned to agriculture. `Service 1' i s trade, transports and services; `service 2' is health, education, andpublic administration. North is Niassa,Cab0 Delgado, Nampula; Center i s Sofala, Tete, Manica, Zambezia; South i s Gaza, Inhambane,Maputo Province,Maputo City. Method. Definepoverty lines I,,i= 1,2 for the two periods. E , withni rows, i s the vector of consumptionof households. W, i s a vector of ones. The poverty rate in period i isp,= (E < l,)' w,/n,. Starting from period i, simulate the poverty rate in penodj , preservingthe distribution of period i but with the averageconsumptionof periodj , by multiplyingall households'consumptionby the ratio of mean consumptionin periodsj and i, thus:pl,= (E( <'?,/n,)i( the change in povertybetween i andj ,p, -p, ,can be written, startingfrom period i, as @,, -p,) + @!, -p,) + residual,where the first ;%,in,) < 23' ;,ini . In the same way, simulatep, . Then element is the growth component and the second the distribution component. Equivalently, starting from periodj , it is @, -po) @, - + pj,)+ residual. Then the two growth componentsare averaged, the two distribution components are averaged, and the two residual components are averaged, yielding the numbers above. 146 ANNEX1: TABLESAND FIGURES Annex Table 17. Decompositionof the change in poverty by geographicaland sectoraldimensions Mozambique North Center South Poverty in 1996 69.1 65.9 73.4 66.1 Poverty in2002 54.1 55.4 45.2 66.8 Total change inpoverty 2002-1996 -15.1 -10.5 -28.2 0.7 Regional decomposition Change inpoverty inthe North -3.4 Change inpoverty inthe Center -12.0 Change inpoverty inthe South 0.2 Total intraregional component -15.2 Population shift (regional migration) -0.03 Interaction component (residual) 0.I Provincial decomposition Change inpoverty inNiassa -1.0 -3.0 Change inpoverty in Cab0 Delgado 0.5 1.5 Change inpoverty inNampula -2.9 -9.1 Change inpoverty in Zambezia -4.6 -11.0 Change inpoverty inTete -1.6 -3.7 Change inpoverty inManica -1.1 -2.6 Change inpoverty in Sofala -4.7 -11.1 Change inpoverty inInhambane -0.2 -0.8 Change inpoverty in Gaza -0.4 -1.5 Change inpoverty inMaputo 0.3 1.2 Change inpoverty inMaputo city 0.4 1.4 Total intraprovincial component -15.4 -10.6 -28.4 0.4 Population shift (provincial migration) 0.02 -0.I -0.1 0.4 Interaction component (residual) 0.3 0.2 0.3 -0.I Urban-rural (consistent 1996 definition) Change inurban poverty -2.5 -7.1 -2.4 2.7 Change inrural poverty -12.7 -4.2 -25.8 -1.9 Total intrasectoral component -15.1 -11.3 -28.2 0.8 Population shift (urban-rural migration) 0.0 -0.6 -0.1 -0.1 Interaction component (residual) -0.0 1.3 0.1 0.0 Aggregate sectors Change inagriculture poverty -11.3 -6.1 -21.9 -0.4 Change inindustry poverty -0.7 0.0 -1.8 1.1 Change inservice1 poverty -0.9 -0.6 -2.3 1.3 Change in service2 poverty -0.8 -1.2 -0.9 0.1 Total intrasectoral component -13.7 -7.9 -26.9 2.I Population shift (sector shift) -1.6 -1.4 -1.0 -1.8 Interaction component (residual) 0.2 -1.2 -0.3 0.4 Individuals are assigned to the sector where the household head i s employed. If the head i s not employed they are assigned to the sector o f employment of the oldest adult. Ifnobody works (less than 5% of all cases) they are assigned to agriculture; `Service 1' includes trade, transports and services; `service 2' includes health, education, and public administrahon. North includes Niassa, Cab0 Delgado, Nampula; Center includes Sofala, Tete, Manica, Zambezia; South includes Gaza ,Inhambane, Maputo Province, Maputo City Method. Define matrices of poverty rates pi,where i is time and the rows @il,p,2,pi3,...)are regions or sectors. Similarly define matrices o f population shares . The nationalpoverty rate inpenodi isp,' c,. The change in poverty rates between periods Iand 2 i s p2' G)is the population shift effect, and c = (p;-6)'(E- E)is the interaction effect. -51`& ( = -15.1% above). This is trivially decomposed as a + b + c, where a = (p;. -.m)'E i . s the intra-sectoral effect,- b ="(E . . -- 147 MOZAMBIQUE COUKTRY ECONOMIC MEMORANDUM Annex Table 18. Householdheademployment sector and contract (1996/7 and 2002/3) All 1'' quintile 2"dquintile 31dquintile 4" quintile 5" quintile Urban Rural 96 02 96 02 96 02 96 02 96 02 96 02 96 02 Agriculture 82 75 87 80 86 80 82 81 84 79 74 60 42 89 Mining 1 1 1 1 1 1 1 0 1 1 1 1 1 1 Manufacturing 4 1 4 1 4 1 3 1 3 1 5 2 2 1 Construction 1 3 1 3 1 3 2 4 2 2 1 3 7 1 Transport 2 2 1 1 1 1 1 1 1 1 2 3 4 0 Trade 4 8 2 5 2 7 5 6 4 7 6 1 4 18 4 Services 3 6 2 6 2 4 2 5 2 4 5 8 15 2 Education 1 2 1 1 1 1 1 1 1 3 2 5 4 2 Health 1 1 1 1 0 1 1 0 1 0 1 1 2 0 Public admin. 2 2 1 1 1 1 2 1 1 2 3 4 6 0 Source: Fox et al. (20051, Table 14, p. 40. 148 ANNEX 1: TABLES AND FIGURES Annex Table 19. Consumption regressions 199617 2002/3 Dep.variable: In consumptionper Urban Rural Urban Rural adult equivalent Coef. Signif. Coef. Sig. Coef. Sig. Coef. sig. HHdemographics No of children 0-5 -0.067 *** -0,051 *** -0.061 *** *** -0.045 *** *** No of children 6-9 -0.146 *** -0.089 *** -0.093 -0.076 No of children 10-14 -0.030 *** ** -0.112 *** -0.106 *** -0.108 *** No ofmen 15-59 -0.081 -0.091 *** -0.003 -0.064 *** No of women 15-59 -0.004 -0.054 *** -0.021 ** -0.028 *** No of adult >60 0.025 -0.089 *** 0.025 -0.028 Head female -0.461 *** -0.122 0.092 -0.186 Any disabledadults -0.171 *** -0.010 -0.052 -0.100 *** Any disabledchildren -0.069 -0.039 0.020 -0.052 Age head 0.024 *** -0.009 *** 0.006 -0.007 ** Age head square -0.000 *** 0.000 *** -0.000 0.000 ** Head marital statusa (base categoyhead single) Head married -0.253 *** -0.142 ** -0.027 -0.141 Headpolygamous -0.192 -0.091 0.001 -0.024 Head divorced -0.242 *** ** -0.001 0.030 -0.058 Head widowed -0.389 -0.113 -0.277 ** -0,127 Added effect offemale head on marital status Head female*married 0.584 *** 0.134 0.102 0.385 ** Head female*polyg 0.285 0.081 0.036 0.199 Head female*divorce 0.390 ** -0.011 -0.209 0.057 Head female*widow 0.632 *** 0.085 0.237 0.171 Head education (base categorq-head no education) Head some education 0.190 *** *** 0.070 *** 0.129 *** 0.062 *** Head completed epl 0.409 0.184 *** *** *** *** *** 0.234 0.131 Head completedep2 0.615 *** 0.159 0.451 0.298 *** Head completedesl 0.712 *** 0.458 *** 0.715 *** 0.695 *** Head over esl 0.996 *** 0.688 *** 1.142 *** 0.542 *** Employment sector (base categoryhead in agriculture) Head mines -0.096 0.276 *** 0.231 *** 0.174 Headmanufacturing 0.156 *** 0.014 0.014 0.275 *** Headconstruction 0.138 0.072 0.036 0.038 *** *** *** *** *** *** ** Head transport 0.22s 0.362 *** *** 0.293 0.660 Head trades 0.343 0.334 0.304 0.296 Head services 0.232 *** 0.372 *** 0.113 *** 0.158 *** Headeducation 0.098 0.256 *** -0.072 0.283 *** Headhealth -0.025 0.292 *** 0.267 *** 0.341 *** Headpublic administr 0.124 *** *** *** ** 0.355 0.156 0.132 Constant 9.016 9.226 *** 9.049 *** 10.174 *** District fixed effectsb Yes yes Yes Yes Observations 2428 5782 4001 4695 Adj Rsq 0.340 0.392 0.364 0.374 *** significant at I%,significant at 5%, ** *significant at 10% Headmarital status:Includinginteraction terms with the gender of the householdhead. See further explanation in Fox et al. (2005). In 1996, 128districts were covered; in2002, 144districts. Source: Fox et al. (2005). 149 MOZAMBIQUEmTRY ECONOMIC C O MEMORANDUM Other income ' 1Income from selling other 4-w Income from selling agric ii: Self-employment income IEmploymentincome Auto-consumption 0 0 c\I all urban rural Annex Figure 20. Sources of cash income (excluding private transfers) and self- consumption, by location, 2002/3179 17' Source: Fox et al. (2005), Graph 4, p. 45. 150 ANNEX 1:TABLESAND FIGURES 100% Occasional income 80% Incomefrom capital 60% lncom fromselling other proc Incomefromselling fish 40% Incomefromselling anim Incomefromselling agric 20% 0Totalself-empl income 0% Total employment incom 1 2 3 4 5 1 Self-consurrption Annex Figure 21. Sources of cash income (excluding privatetransfers) and self- consumption,by quintile,20O2/3l8O '*'Source:Fox et al. (2009, Graph4, p. 45. 151 a2 d z0 d0 0 E a, d, ir 0 B6 Y I a, 3 ir 0 :: YV % Ph nc 0 .I Ya E v1 1 e *8 *ma Ia, e *- r I Y 'J 1 m L a, a c 0 Bm v) v P 4- zm L 0L :$ 4 N a, .-R L a p1, 42 3 ANNEX1:TABLES AND FIGURES Annex Table 23. Employmentby sector, by rurayurbanlocation and by poorlnon-poor 1996197 2002103 Rural Urban All Rural Urban All P N P P N P P N P P N P P N P P N P Ag., for., fish. 96.0 93.3 68.5 49.6 91.6 84.1 92.6 86.8 47.3 31.6 85.1 75.9 Mines 0.3 0.5 1.3 1.0 0.4 0.6 0.3 0.6 0.8 0.9 0.4 0.6 Manufacturing 1.2 1.2 5.6 8.6 1.9 2.8 0.3 0.6 1.9 2.0 0.6 0.9 Construction 0.5 0.5 2.9 3.2 0.8 1.1 1.4 1.1 6.7 4.6 2.3 1.8 Transport 0.2 0.3 2.5 5.0 0.6 1.3 0.3 0.7 2.3 4.5 0.6 1.4 Trade 0.8 1.7 9.3 14.5 2.2 4.4 3.0 5.1 19.7 22.7 5.8 8.5 Services 0.4 1.0 5.7 9.0 1.3 2.7 1.4 2.3 16.1 22.0 3.8 6.2 Education 0.3 0.8 0.9 2.5 0.4 1.2 0.4 2.0 1.9 4.4 0.6 2.5 Health 0.2 0.3 1.0 1.6 0.3 0.6 0.2 0.4 0.4 1.9 0.2 0.7 Public admin. 0.2 0.4 2.3 5.2 0.5 1.4 0.2 0.6 2.8 5.4 0.6 1.6 Source: Fox et ai. (2005 ntum from IAF2002!3. h'ote that this table inclu s dornisricos in the agriculturallabor force. Annex Table 24. Sector employment rates, by sex and urbanhral (percentage) 1996 2002 2002 All Men Women All M e n Women Urban Rural Agriculture 89.2 80.7 95.9 80.7 68.7 90.1 51.3 93.0 Mines 0.5 1.1 0.0 0,5 1.o 0.1 0.9 0.4 Manufacturing 2.2 4.2 0.6 0.7 1.5 0.1 1.6 0.4 Construction 0.9 2.0 0.1 2.0 4.5 0.1 4.7 0.9 Transport 0.8 1.7 0.1 1.o 2.2 0.1 2.9 0.2 Trade 2.9 4.1 1.9 7.1 9.6 5.1 17.2 2.9 Services 1.7 3.1 0.6 5.0 7.4 3.1 14.1 1.1 Education 0.7 1.1 0.3 1.5 2.6 0.7 3.0 0.9 Health 0.4 0.5 0.4 0.4 0.5 0.4 1.1 0.2 Public administration 0.8 1.6 0.2 1.1 2.1 0.3 3.2 0.2 All 100 100 100 100 100 100 100 100 Sample: all those aged 10-59 and working. Source: Fox et ai. (2005), Table 19,p,46. 153 MOZAMBIQUE COUNTRY ECONOMICMEMORANDUM Annex Table 25. Type of contract, by industryand by urbanhuralin 2002 (men andwomen) Receiving Casual Family Self- a wage workers worker employed All p e r c e n t a g e Agriculture 1.7 0.5 43.9 53.9 100 Mines 59.5 5.4 0.9 34.3 100 Manufacturing 71.5 6.8 1.6 20.1 100 Construction 50.1 37.2 0.8 11.8 100 Transport 75.4 11.1 2.0 11.6 100 Trade 19.7 2.8 7.3 70.3 100 Services 70.7 8.5 13.0 7.8 100 Education 99.1 0.0 0.9 0.0 100 Health 85.5 0.9 0.7 13.0 100 Public administration 100.0 0.0 0.0 0.0 100 Urban 30.2 5.3 23.2 41.3 100 Rural 3.9 0.6 44.9 50.6 100 All 16.1 2.6 33.7 47.6 100 Sample: all those aged 10-59 and working. Source: Fox et al. (2005) Annex Table 26. Type of contract, by industryand by urbanhral in 2002 (men) Receiving Casual Family Self- a wage workers worker employed All p e r c e n t a g e Agriculture 3.2 0.8 22.0 74.0 100 Mines 60.1 5.9 0.0 34.0 100 Manufacturing 71.5 7.4 1.5 19.6 100 Construction 49.4 37.7 0.8 12.1 100 Transport 74.9 11.6 1.8 11.7 100 Trade 22.3 3.7 5.2 68.8 100 Services 75.9 10.8 4.2 9.2 100 Education 99.9 0.0 0.1 0.0 100 Health 82.8 1.8 1.4 14.0 100 Public administration 100.0 0.0 0.0 0.0 100 Urban 45.2 9.4 10.9 34.6 100 Rural 7.8 1.2 19.4 71.6 100 All 23.4 4.3 15.1 57.2 100 Sample:all those aged 10-59and working. Source: Fox et al. (2005) 154 AWEX1:TABLES AND FIGURES Annex Table 27. Type of contract, by industryandby urbadruralin 2002 (women) Receiving Casual Family Self- a wage workers worker employed All p e r c e n t a g e Agriculture 0.3 0.2 64.5 35.1 100 Mines 52.5 0.0 10.0 37.5 100 Manufacturing 72.0 0.0 2.6 25.5 100 Construction 88.1 11.9 0.0 0.0 100 Transport 85.2 0.0 6.2 8.6 100 Trade 15.2 1.3 10.7 72.8 100 Services 60.1 3.7 31.2 5.1 100 Education 96.4 0.0 3.6 0.0 100 Health 88.1 0.0 0.0 12.0 100 Public administration 100.0 0.0 0.0 0.0 100 Urban 14.3 1.o 36.3 48.4 100 Rural 0.9 0.2 65.O 34.0 100 All 7.1 0.5 56.7 35.8 100 Sample: all those aged 10-59 and working. Source: Fox et al. (2005). ,Vole: There i s no fully comparablequestion in the 1996197 IAF survey. In 1996197 no distinction was madebetween receiving awage and being a casual worker. When comparingthe self-employed category(which was a separatecategory inboth surveys) the total percentage does not change over both years. The percentagefor men decreasesby 0.1% and for women increasesby 0.7%. Annex Table 28. Type of contract, by industry,locationand poorhon-poor, 200213 POOR NON-POOR Wage Casual Family Self- All Wage Casual Family Self- All earner worker worker empl. earner worker worker empl. Ag., for., fish. 1.4 0.7 51.9 46.0 100 1.5 0.1 46.1 52.3 100 Mines 71.0 1.9 6.5 20.6 100 57.3 6.3 2.2 34.2 100 Manufacturing 67.5 9.1 1.3 22.1 100 72.5 4.0 1.5 22.0 100 Construction 44.2 42.6 0.8 12.4 100 58.5 27.5 0.4 13.5 100 Transport 77.8 20.8 1.5 100 73.4 8.5 2.6 15.5 100 Trade 16.5 3.9 6.1 73.5 100 19.9 3.7 8.6 67.8 100 Services 76.9 8.5 5.2 9.4 100 65.4 7.7 20.1 6.8 100 Education 100.0 100 97.7 2.3 100 Health 64.6 5.5 2.1 27.8 100 86.8 2.3 11.0 100 Public admin. 100.0 100 100.0 100 Urban 25.1 6.6 27.4 40.8 100 38.7 4.6 20.3 36.5 100 Rural 3.3 0.9 50.3 45.4 100 5.2 0.4 43.6 50.8 100 All 8.9 2.4 44.5 44.3 100 15.0 1.6 36.8 46.6 100 Source: Fox el al. (20( I. in turn from IAF2002/3. Note that this table in( des domdsticos in the agricultural labor force. 155 MOZAMBIQUE COUNTRY ECOKOMIC MEMORANDUM Annex Table 29. Comparativelabor regulationsandworker benefits Mozambique South Africa Lesotho Namibia Botswana Vormal 48 per week 45 per week 45 per week 45 per week 48 per week working hoursa Overtime <8 p.m. = 1 . 5 ~ Standard= 1Sx Standard= 1.25~ Standard= 1 . 5 ~ Standard = 1 . 5 ~ Pay 28 p.m. = 2 . 0 ~ ' Sunday= 2 . 0 ~ Holidays = 2 . 0 ~ Sunday = 2 . 0 ~ Sunday=2 . 0 ~ Max = 2 hrs/day Max = 10hrs/wk Max = 11hrsiwk Max = 10hrs/wk Max = 14hrshvk Vacation Yr 1 =21 days Yr 1 = 21 days 1dayione-month worked 24 days Yr 1= 15 days >Yr 2 = 30 days, not > Yr 2 = 1 dayil? days >Yr 2= 1.25 daydone- including public holidays worked month worked & sick leave Sick leave No limit <6 months worked = 1 <6 months worked = zero >3 years worked = 30 14days full pay Medical certificate day126 days worked >6 months worked = 12 days Medical certificate provided after 30 days Every 3 years = #of days days provided worked in 6 weeks >I2 months worked = 12 Medical certificate days full pay, 24 days provided halfpay Maternity Pre-birth= 20 days Max = 4 months unpaid Pre-birth= 6 wks Pre-birth= 4 wks Pre-birth= 6 wks leave Post-birth= 40 days Post-birth= 6 wks Post-birth= 4 wks Post-birth= 6 wks Full pay Unpaid leave Unpaidleave 25% of wage Dismissal 3 mths <6 mths worked = 1 wk <6 mths worked = 1 wk <4 wks worked = 1 day <5 yrs worked = variable notice >6 mths worked = 2 wks >6 mths worked = 2 wks <12 mths worked = 1 5-10 yrs worked = 4 wks >Imthsworked=4 2 >I2 mths worked = 1 wk wks mth >I2 mths worked = 4 wks Severance 3-6 mths worked = 45 1 yr worked = 1 wk Iyrworked=1wk I)Tworked=1wk >5 yrs worked begin pay days accrual 6 mths- 3 yrs worked = 3 mths >3 yrs worked = 3 mths for every 2 yrs worked Source: SAL Consultoriae Investimentos, LDA (2002), Annex 2. Normal working hours as defined by law. In Mozambican law: Law No. 8/98, article 28: 1. In Mozambique:up to a maximum of 100hours of overtimeper year. Law No. 8/98,article 33:3. In Mozambique:overtime pay- in Law KO.8/98, Article 48:l. 156 ANNEX 1: TABLESAND FIGURES 40 35 Moz 0 Kenya 15 0 Zambia 0 Tanzania 5 4 Uganda 0 Ethiopia 0 50 100 150 Firing costs (weeks'wages) . Ann Table 30. Firingcosts and firm perceptions insub-Sahar n Africa'81 40 35 0 Mozamb. 0 Kenya 15 0 Zambia 0 Tanzania 5-1 O Ethiopia I .,I 0 20 40 60 80 Difficulty of firing index Annex Table 31. Difficultyof firing and firm perceptions in sub-Saharan Africa'*' 18' Firing costs from World Bank, Doing Business in 2004. See hth,:/;rru.worldbank.or~/Doiii~Business/Exr,loreTo~ics/HiringFiriii~~Workers/ComuareAll .asux?direction- asc&soh-2 , Firmperceptions o f labor as a problem: from Alby et al. (2005). Intum from the WED or ICAs. Indicates the percentage o f firms which have associated an obstacle index o f 3 or 4 to each issue, where 0 = no obstacle, 1 = minor obstacle, 2 = moderate obstacle, 3 = major obstacle, and 4 = very severe obstacle. The firmperception o f labor as a problem for Mozambique was taken from World Bank (2003a), Table 2.33, p. 36, referring to the opinion about "Layoff procedures and cost o f retrenchment". 18* Difficulty o f hiring index from World Bank, Doing Business in 2004. See http:linu.worldbank.orp/DoineBusinessiExuloreTouicsiHirinrzFiringWorkersiComuarcAll.asux'!direction= asc&sort=2 , Firmperceptions o f labor as a problem: from Alby et al. (2005). Intum from the WED or ICAs. Indicates the percentage o f firms which have associated an obstacle index o f 3 or 4 to each issue, where 0 = no obstacle, 1= minor obstacle, 2 = moderate obstacle, 3 = major obstacle, and 4 = very severe obstacle. The firmperception o f labor as a problem for Mozambique was taken from World Bank (2003a), Table 2.33, p. 36, referringto the opinion about "Layoff procedures and cost o f retrenchment". 157 MOZAMBIQUE COUNTRYECONOMIC MEMORANDUM Annex Table 32. Wage regressions,includingagricultural workers, 2002/3 Dependentvariable: Inweekly All Men Women Sig of wage Coeff. Sig Coeff Sig Coeff. Sig difference Age 0.060 *** 0.063 *** 0.055 *** Age squared -0.001 *** -0.001 *** -0.001 ** Female(1=Q -0.280 *** Marital status Married 0.254 *** 0.329 *** 0.137 Polygamous 0.215 *** *** 0.267 *** -0.029 Cohabiting 0.117 0.155 *** 0.142 * Divorced 0.095 0.153 ** * 0.081 Widowed 0.113 0.326 0.154 Education CompletedEP1 0.221 *** 0.206 *** 0.224 *** CompletedEP2 0.5 10 *** 0.473 *** 0.666 *** * CompletedES1 0.842 *** 0.786 *** 1.074 *** ** CompletedES2 1.132 *** 1.066 *** 1.412 *** ** CompletedET1 1.001 *** 0.929 *** 1.235 *** CompletedET2 1.484 *** 1.406 *** 1.605 *** Teacher education 0.929 *** 0.961 *** 1.086 *** Higher education 2.412 *** 2.333 *** 2.636 *** Industrial sector Agriculture -0.532 *** -0.488 *** -0.630 *** Mining 0.768 *** 0.783 *** 0.342 Construction 0.132 0.113 0.836 *** ** Transport 0.182 *** 0.191 ** -0.030 Trade -0.095 -0.142 0.004 Services -0.031 -0.023 -0.038 Education 0.096 0.098 -0.012 Health 0.098 0.03 1 0.164 Public administration 0.044 0.068 -0.155 Type contract Casual -0.298 *** -0.255 *** -0.760 *** ** Constant 10.366 *** 10.298 *** 10.126 *** District effects Yes Yes Yes Observations 2810 2218 592 Adj Rsq 0.559 0.532 0.665 ***significant at 1%, ** significant at S%, *significant at 10% The dependent variable is the logarithm of the weekly wage in the mainjob (including fringe benefits). Source: Fox et al. (2005). 158 ANNEX1:TABLES AND FIGURES Annex Table 33. Commodity compositionof imports, 1997-2002 (US$ millions) ProductGroup 1997 1998 1999 2000 2001 2002 ~~~~~ ~~~ ~~~ ~~~ ~~ Fish 3.9 9.6 7 8.7 7.8 16.3 Agriculture (excluding fish) 165.6 191 131.2 156.8 143.6 123.3 Rice 40 52 40.6 31.4 45.9 42.2 Wheat 30 30.5 22.2 16.9 25.3 26.4 Maize and maize flour 1.6 6.4 2.9 4.4 4.4 5.7 Vegetable oil 14.2 13.4 20.5 12.4 12.4 1.5 Sugar 30.4 25.1 13.8 12.2 12.2 4.1 Fruitsand vegetables 4.9 5.8 2.9 6.3 6.3 1 Processed food products 8.7 12.7 6.5 17.4 17.9 7.6 Beverages 11.1 10.5 5.2 15.9 9.2 3.9 Minerals and fuel 99.5 85.3 80.3 190.2 183.6 160.5 Cement 6.5 11.8 5.2 6.8 12.4 10.1 Fuels 77.2 59 72.2 158.8 169.7 149.6 Electricity 15 8.4 14.1 14.6 28.5 30.7 Pharmaceuticals 12.9 7.4 8.5 14.5 19.3 24.8 Fertilizers 4.3 2.4 3.2 2.2 4 2.3 Other chemcal products 38.5 41.6 28.3 59.9 78.3 87.2 Soap and detergents 4.4 6 3.5 6.9 6.4 6.7 Insecticides and fungicides 7.5 10.1 5.6 12.6 4.5 9.1 Plastics 13.7 13.2 8.2 17.6 50.3 53.1 Rubber products 12.7 8.4 6.1 16.4 9.5 10.1 Wood and paper products 16.2 25.7 10.5 30.1 40.1 26.5 Textiles and clothing 42.3 42.6 29.9 41.5 26.2 19.7 Textiles 19.2 18.9 11.4 17.3 12 18.8 Clothing 5.6 6.4 9.2 6.7 4.1 3.3 Used clothing 12.7 12.6 6.2 10.3 8.8 5.3 Iron and steel 27.2 94.3 26.2 68.9 39.7 50.6 Other metals 9.1 7.9 4.4 10.6 234.7 415.2 Machinery 74.4 85.9 60.5 188.9 131.7 136.1 Transportation equipment 113.4 73.8 174.7 174.8 80.5 135.5 Scientific equipment 12.4 18.6 7.9 23.6 10.7 13.8 Furniture 9.1 0.8 7.8 17.1 7.3 11.7 Other 29.7 n 15.6 26 28.1 26.5 T O T A L 755.9 817.3 656.5 1162.3 1063.4 1262.9 Sources: DTIS, 2005, Table 1-6. In turn from NE, customs bulletins, IMOPETRO, EdM, Mozal and MOTRACO. 159 MOZAMBIQUECOUNTRY ECONOMIC MEMORAKDUM -5 80 - 0 Nepal C -8 70- 0 Bhutan Bolivia Y d= 10- 0 Cambodia 2 0 - Annex Figure34. Accessto bank loansby manufacturingfirms, ICA co~ntries"~ 45 - 3 40 - Bolivia 0 c3 2 35 - Honduras 0 3 Nepal 0 Eritrea 30- sm %angladesh z , 2 5 - 0 Ethiopia 0 Pakistan E 2 . 0 - 'Kenya 0 - .-m E Mozambique 0 1 5 - 0 Nigeria Bhutan Q Tanzania 0 m 1 0 - Uganda EI Zambia ; 5 - Cambodia O i Annex Figure35. Claims on the privatesector as a fraction of GDP, international comparison,2 0 0 1 ~ ~ ~ Sources: Investment Climate Assessments and associated Excel tables, at http://nu.worldbank,org/lnvestmentClimate/. By "bank loans" here is mean termloans, excluding overdraft facilities. 184 Source: IMF,International Financial Statistics, paper version for 2005 March. C l a i m on the private sector: variable 32d. GDP: variable 99b. 160 ANNEX1:TABLESAND FIGURES 350 0 Zambia -c8 300- % 250- A 2 Y- sO 200- Kenya Bolivia o VI Mozambique o 150- .- e! 0 0 Tanzania -E $ loo Ethiopia 0 uga#a 8angladesh 0 China Eritrea 0 Pakistan .$ -- 50- o O" Cambodia 0 1 Annex Figure36. Value of collateralrequiredfor loans, international comparis~n'~~ 18' Sources: Collateral required i s from: Investment Climate Assessments and associated Excel tables, at http:l/rru.worldbank.org/InvestmentClimate/, GDP per capita is from WorldDevelopment Report 2003. 161 MOZAMBIQUE COUNTRYECONOMICMEMORANDUM ANNEX 2 BACKGROUND PAPERS FOR THE CEM Achcar, Asmara, 2004. Poverty reduction in rural Mozambique: Evidencefrom physical measures of output & non-income measures of welfare: 1996-2002. Mimeo, World Bank, February 20. Supported by the Belgian Poverty Reduction Partnership (BPRP). Benito-Spinetto, Maria Teresa, and Peter G. Moll, 2005. Macroeconomic developments, economic growth and consequencesfor poverty. (Background paper for the Country Economic Memorandum). February 24,2005. de Barros, Gilberto, 2005. Private sector development contribution to the Mozambique Country Economic Memorandum. AFTPS. World Bank. de Sa, Paulo, 2004. Backgroundpaper on mining for the Mozambique Country Economic Memorandum. September. World Bank. Eide, Arne, 2004. An economic analysis of natural resourcessustainability in Mozambiquejisheries. August 13. Mimeo, World Bank. Backgroundpaper for the MozambiqueCountry Economic Memorandum. Fox, Louise, ElenaBardasi and Katleen Van Den Broeck, Evolution ofpoverty and inequality in Mozambique, 1996/7-2002/3. (Background paper on poverty for the Country Economic Memorandum). February 17,2005. Hughes, Gordon, 2005. Rural land issues andpolicies in Mozambique. Background paper for the Mozambique Country Economic Memorandum. Hughes, Gordon, 2005. Water issues andpolicies in Mozambique. Backgroundpaper for the Mozambique Country Economic Memorandum. Klasen, Stephan, and Silke Woltennann, 2004. The impact of demographic dynamics on economic development,poverty and inequality in Mozambique. Mimeo, World Bank, September 27. Background paper for the Mozambique Country Economic Memorandum. Supportedby the BelgianPoverty Reduction Partnership (BPRP). Margulis, SCrgio, 2005. An economic analysis of natural resource sustainability in Mozambique. Backgroundpaper for the Mozambique Country Economic Memorandum. April 5, 2005. World Bank. Seroa da Motta, Ronaldo, 2004. An economic evaluation of forestry regulation in Mozambique. November. Background paper for the Mozambique Country Economic Memorandum. 162