Report No. 8835-MAG Madagascar Urban Sector Review June 191 Infrastructure Operations Division South-Central and Indian Ocean Department Africa Region FOR OFFICIAL USE ONLY ..~~~~~~~~~~~~~~~~~~~~V ocumment of. e wodr B-. 'This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be $islos%iwithout World Bank authorization. .5 Currg!ncX Units Malagasy Franc Conversions Based on Exchange Rate (PMGIUS$), 1989: 1,560 FMG 10,000 - US$ 6.40 FMG 1 million = US$ 640 US$ 1 = FMG 1,560 US$ 100 - FMG 156,000 US$ 1 million - FMG 1.56 billion Exchange Rate (FMG/US$), 1991: 1,850 Weights and Measures: Metric System ACRONYMS AND ABBREVIATIONS APU Land Development Agency (Agence fonciere urbaine) ANTAFITA Bus company owned by city of Antananarivo BCT Transport and Traffic Bureau (Bureau de circulation et de transports) BDE National Data Bank (Banque des donnees de l'Etat) BFV Commercial bank (commerce) BNI Commercial bank (industry) BPU Urban Projects Bureau (Bureau des projets urbains) BTM Commercial bank (agriculture) CD Local Government (Collectivite decentralisee) CNAPS National savings fund (Caisse nationale de prevoyance sociale) CNC National Coordinating Committee (Comite national de coordination) CTNDU National Technical Committee for Urban Development (Comite technique national pour le developpement urbain) CTTUS Technical Committee on Urban and Suburban Transport (Comite technique pour le transport urbain et suburbain) DAUH Directorate of Architecture, Urban Planning, and Housing (Direction de l'architecture, de l'urbanisme, et de l'habitat) DEPCT Directorate of Studies, Programming, and Coordination of Transport (Direction des etudes, de la programmation, et de la coordination des transports) DP Property Directorate (Direction du patrimoine) EDF European Development Fund EEM Eau et electricite de Madagascar (former Antananarivo municipal corporation) FAC French aid agency (Fonds d'aide et de cooperation) FAR Fivondronana of Antananarivo-Renivohitra (Municipality of Antananarivo) FIBATA Bus company, Antananarivo FOR OFFICLAL US ONLY FIHU Fund for Infrastructure and Urban Housing (Fonds d'infra- structure pour l'habitat urbain) FIMA Bus company, Antananarivo FMG Malagasy Franc (Franc malgache) FNDE National Economic Development Fund (Fonds national de developpement economique) GDP Gross Domestic Product IDA International Development Association (World Bank) IFL Local-purpose investments (Investissements a fonction locale) IMF International Monetary Fund JIRAMA Malagasy Water and Electricity Authority (Jiro sy Rano Malagasy) Kwh Kilowatt-hours LTIM Long-Term Image of Madagascar LTPS Long-Term Perspectives Study Ml Ministry of Interior (Ministere de l'interieur) MIEL Mutual savings society (Mutuelle inter-epargnants pour le logement) MIEM Ministry of Industry, Energy, and Mines (Ministare de l'industrie, de l'energie, et des mines) MINEP Ministry of Economy and Planning (Ministere de l'economie et du plan) MPAPF Ministry of Agricultural Production and Land Resources (Ministare de la production agricole et du patrimoine foncier) MTMT Ministry of Transport, Meteorology, and Tourism (Ministere des transports, de la metorologie, et du tourisme) MTP Ministry of Public Works (Ministere des travaux publics) Mwh Megawatt-hours PCE President of Executive CommitteelPrsident du comite ex&eutif (Mayor) PDUM Firat Urban Project, Madagascar (Projet de developpement urbain, Madagascar) PEP Public Expinditure Program PER Public Expenditure Review PME Small and medium-scale enterprises (Petites et moyennes entreprises) PUF Urban Project in the Faritany (Projet urbain dans les Faritany) SAM Social Accounting Matrix SCD Direct Tax Bureau (Service des contributions directes) SD Land Registry Bureau (Service des domaines) SE.TMAD National Housing Authority (Societe d'equipement immobilier de Madagascar) SP MIEL Firm for promotion of mutual savings (Societe de promotion MIEL) SRPCH Regional Bureau for Roads and Bridges (Service regional des ponts et chaussees) SSA Sub-Saharan Africa TAFB Additional Tax on Built Property (Taxe annexe a l'impot foncier sur la propriete batie) TUr Transactions Tax (Taxe unique sur les transactions) UNDP United Nations Development Programme This document biss a restricted distribution and may be used by recipients only in tho performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorIzation. This report is based on the findings of a mission to Madagascar In October 1989, and of subsequent visits by several tea members. The study was undertaken at th request of the Ministry of Economy and Planni (HINEP), which ensured the coordiation of its activities. The issues of special focus during the mission were proposed by the National Technical Committee for Urban Development (CTNDU), which, together with HINEP, supervised the mission's work. The mission was composed of Orville Grtimes (mission chief), Michel Arnaud (urban specialist, consultant), Mohammed Benouahi (financial analyst), Jean-Marie Cour (urban economistIdemographic specialist), Gilles Horoenfold (housing finance specialist, consultant), and Tidjane Thiam (local finance specialist). Contributions to the report wero prepared by Michel Arnaud, Jean-Marie Cour, Catherine Farvacque, Orville Grimes, Gilles Norenfeld, and Tidjan. Thiam with Mohammed Benouahi. Peter Hasell provided helpful advice on findings of the SAM model. The report wa discussed with the Governmt authorities concerned in February 1991. Report processing and presentation were the responsibility of Catherine utie, Marie-Marthe J;rsmr, Armelle Nau, Connie Kok Shun, and Orville Grimes. MADAGASCAR URBAN SECTOR REVIEW Table of Contents Pame No. SUMMARY OF CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . . . I PART I: THE ECONOMIC SETTING OF URBANIZATION Chapter I: The Real Econoum of Madagascar In an Urban-Rural Perspective .......... . . . . 1 A. Introduction . . . . . . . . . # .. * . . .. . * 1 B. National Resources and Expenditures in 1988 . . . . . . . 2 Social Accounting Matrix (SAM) of the National Economy . 8 C. The Rural Economy and the Agricultural Sector . . . , . . 8 D. The Urban Economy . . .. ........ . . . . 10 E. Urban-Rural Exchanges . . . . . . . .......... 11 F. The Case of Antananarivo . . . . . ...... ... 17 Chapter IIs Urban Policies in the 1980s 9 . . . . . . . . . . . 21 A. Characteristics of Urban Growth . . . . . . . . . . . . 21 Squatter Settlements . . . . . 25 Water and Electricity Networks ............ 26 B. Sector Performances Battling the Odds . . . . . . . . . 27 Housing * a . . 9 9 . 9 9 . . 9 . 9 . 9 9 . . . a . 28 Urban Planning . . . . 9 9 . . . . . . . . . . 29 Operations . . . . . . . . . . . . . . . . . . . . . . 29 Urban Land Management . . . . . . . . . . . . . . . . . 31 Water and Electricity . . . . . . . . . . . . . . . . 31 MunicipalServices 99. 999.999.99.32 PART II: A GROWTH-ORIENTED URBAN STRATEGY Chapter IIIs Medium-Term Scenarios of Growth and Urbanization . . . 33 A. Introduction . . . . . * . . . . . . . . . . . . 33 B. Internal Migration and Urbanization Policies . . . . . . 33 C. The Sustainability of Urbanization . . . . . . . . . . . 34 D. Medium-Term Scenarios of Urbanization and Economic Growth . . . . . a * * 9 9 9 9 9 9 9 * . 35 Chapter V: Urban Expenditure Programs ... 3.9.... ... 59 A. Urban Expenditure Programing In the 19809 . . . . . . . 39 Public Expenditures in the Urban Sector . . . . . . . . 39 S. Tovard an Urban Investment Program . ... .. ... . 41 A Consistency Framvork for Planning ......... 41 Elements of a 1991-1995 PEP for Urban Areas . . . . . . 43 PART IIIs STRATEGY IMPLEMENTATION Chapter Vs Strenzthenina Urban Institutions . . . . . . . . . . . . 47 A. National Technical Committee for Urban Development (CTNDU) . . . . . . . . . . . . . . . . . . . . . . . . 47 B. Decentralization and Local Government . . . . . . . . . 48 Legal Status and Organization. . * * ........... * . 48 Personnel and Staffing .44.4.. 4* 44* 44* 49 Financial Resoures for CDs . . . . . .. . .. . .. . SO Building the City. . .............. . . . 50 C. The Nanag.ment of Antananarivo . . . . . . . . . . . . . 51 D. Ministry of Public Works (MTP) . . . . . . . .*. . . . . 53 Organization . . . . 4 . . . . . . . . . . 53 Project Coordination and Management . . . . . . . . . . 54 E. Ministry of Interior ... ............... 55 F. Selmad S * . . * 4 4 4 4 4 56 G. Jirama S* . . . . . . 4 4 57 H. TrainingRequirements .,44.... 44 . . 57 Chapter VTs Local Government Finance and Financial Manaatement . . 59 Local Governent Finance .... . . ... 59 A. Central Government Transfers . . . . . . . . . . . . . . 61 B. Local Taxes 6 * * . . . * * o . 4 4 * * * * * 63 Department of Taxation 4*44*4444 64 1. Tax base determination . . . . . . . . . . . . . . . 64 2. Tax computation 65 35 Tax registers 66 Treasury Department. . . . . . . . . . . . . . . . . 67 1. Tax collection . . . . . . . . . . . * * * 67 2. Bad debts . . . . . . . . . . . . . . . . . . . . . 68 3. Treasury department budgets 4 . . . . . . . . . . 4 69 C. Other Revenues . . .. . . . . . . . . . . . . . . . . 69 1. Revenues from Goods and Services . .. o... . . . . . 69 2. Parafiscal Revenues . . . . . . . .. a. . . . ... . 70 Financial Management . . . . . 4 * 4 * * * * . . . .71 A. Legal Framework and Budgetary Procedures . . . . . . . . 71 B. Expenditure Programming and Control . . . . . . . . . .72 1. Current Expenditures os......... . 73 2. Capital Expenditures. .4. 94444444 77 Chapter VIIs Urban Land Manaaeunt . . . . . . . . . . . . . . . . 80 A. tand for Urban Growths Concept and Reality . . . . . . . 80 Setting . . . . . . . . . . . . . . . . . so ...8 Land Rights and the Registration System . . . . . . . . 81 Development Control and Guidance . . . . . . . . . . . 82 Settlement of Urban Land . . . . . . . . . . ... 83 B. Options for Improvement . . . . . . . . . . . . . . . . 84 Objectives . . . . . . . . . . . . . . . . * . . . . . 84 Improving Land Development Procedures . . . . . . . . . 84 Iacreasing the Supply of Developable Land . . . . . . . 86 C. ImplemntationIssues ................. 86 Chapter Vllls Housint and Housin.PinFance. . . . . . . . . . . . . . 90 A. The Foundations of Housing Policies . . . . . . . . . . 90 Urbanization, Public Investment, and Housing Finance . 90 Savings and Investment in Rousing . . . . . . . . . . . 91 Access to Property and Rental Housing . . . . . . . . . 92 B. Savings Mobilisstin and the Financing of Housing . . . 94 Mutual Associations and Cooperatives . . . . . . . . . 95 The Banking System and Rousing Finance . . . . . . . . 96 Savings-for-Housing Programs . . . . . . .. . . . . . 98 Other Resources of the Banking System . . . . . . . . . 99 A HousingFinance Institution? . . . . . . . . . . . .101 Employee and Civil Servant Housing . . . . . . .. .. 101 C. Summary of Recomendations . . . .. . ... . . . . 102 Chapter IXs Urban Transport . . . . . . . . . . . . . . . * . . 104 A., Introduction . . . . . . . . . . . . . * * . . . * . . .104 B. The Public Trnsport Crisis In Antananarivo . . . . . . 104 Transport Operators . . . . . . . . . . . . . . . . . 104 Users . . . . . * . 9 . . . . 105 Fares . . . . . . . 106 Road Network . . . . . . . . . .. . . . . . . . . . 106 Sector Organzdation and Management . . . . . . . . . . 107 C. Recovery Measures for Public Transport . . . . . . . . . O107 List of Table. Table 1.1 Distribution of Total Population in 1988 . . . . . . . . . 4 Table 1.2 Gross Domestic Product by Sector . . . . . . . . . . . . . 5 Table 1.3 Urban and Rural Contributions to GDP . . . . . . . . . . . 6 Table 1.4 Total Househol -,xpenditure in 1988 . . . . . . . . . . . 7 Table 1.5 Household Expen .:.;ure on Imported Goods, Year 1988 . . . . 7 Table 1.6 Social Accounting Matric of the Malagasy Economy in 1988 . . 00. *0e*000. *t0*. .................................. 9 Table 1.7 Social Accounting Matrix of the Urban Economy of Madagascar In 1988t Intra-Urban Transactions, and Transactions between Urban Areas, Rural Areas and the Rest of the World .. . .. .. .. . ,..... 13 Table 1.8 Social Accounting Matrix of the Urban Economy of Madagascar in 1988: Detail of Receipts from the Rural Areas and from the Rest of the World. ......... 14 Table 1.9 Macro Aggregates of Urban and Rural Areas, Year 1988 . . 17 Table 1.10 Macro Aggregates of the Antananarivo Region, Year 1988 . 18 Table 1.11 Household Expenditure of the Antananarivo Region, Year 1988 . . . . . . . . . . . . . . . . . . . . . 20 Table 2.1 Urban Population Growth . . . . . . . . . . . . . . . . 24 Table 3.1 Madagascar: Two Scenarios of Population Redistribution and Economic G!-Ywth . . . . . . . . . . . . . . . . . . 36 Table 4.1 Projection of Local-Purpose Investment ('IFL") for 1993 45 Table 4.2 Projection of "IFL' Investment for 1993: Allocation for Expansion of Urban Areas and for Rehabilitation . . 46 List of Annexes Annex 1.1 An Accounting Framework for the Real Economy (SAM)s Main Features. . . . . * . * . a . . . . . . . . . . . . 110 Annex 2.1 Population of Former Urban Communes of Fivondronana and Firsisana . . . . . . . . . . . . . . . . . . . . . . . 115 Annex 4.1 Preliminary Ideas on Preparation of a Long-Term Image for Madagascar . . . . . . . . . . . . . . . . . . . . 116 Annex 5.1 List of Former Urban Communes . . . . . . . . . . . . . 121 Annex 6.1 Local Governments . . . . . . . . . . . . . . . . . . . 124 Annex 6.2 Accounts and Budgets . . . . . .......... . . . 126 Annex 6.3 Finances of Former Urban Communes . . . . . . . . . . . 127 Annex 6.4 Analysis of Budgets of Faritany . . . . . . . . . . . . 130 Annex 6.5 Analysis of Budgets of Fivondronana which are former urban communes . . . . . . . . . . . . .,. .*. . 143 Annex 6.6 Analysis of Budgets of Flralsona which are former urban communes . . . . . . . . . . . . . . . . . . . . .. 150 Annex 6.7 Greater Antananarivot Size Distribution of Land Plots . 157 Annex 6.8 FIHU Decree (Decret No. 87-287 portant creation du Fonds de l'infrastructure pour lVhabitat urbain (F.I.H.U.)) . 158 Annex 8.1 The Urban Housing Market . . . . . . . . . . . . . . . . 160 Annex 8.2 Contractual Savings for Housings Illustrations . . . . . 165 Annex 8.3 The MIEL Projects Evolution and current status . . . . . 166 M NP IBRD No. 20035R SULMfY OF CONC2LUSIONS AND RECOMNDATIONS 1. The 1980. seemed to confirm to the Malagasy authorities that they had Irretrievably lost the struggle to manage urbanisation. The growth of the urban population, although moderate by standards both of other countries and of Madagascar In the 19609 and early 1970., was hardly matched by any Investment in infrastructure or services. Buildable land, credits for housing, and transport were available to the top Income groups of cities and towns, but to almost no others. Local governments were charged with wide-ranging responsibilities but had few resources with which to discharge them. The capital of cities was bSing drawn down without rehabilitation or replacement. 2. Two parallel events of the late 1980s began to alter this outlook. The first was the increaseJ awareness that, in addition to the external sector, the domestic economy had the potential to contribute significantly to growth. Urban->e.sed activities are responsible for 45X of the GDP, and can contribute even -,. re to growth through the Intensification of rural-urban exchanges. This in tern requires more effective management of urban areas, by local authorities, the central government, and the private sector in partnershlp. 3. Second, practical experience was gained with policies that had previously been understood only at the co.nceptual level. The city of Toamasina, especially, demonstrated that the restructuring of rather disorderly squatter settlements was technically possible, that land tenure problems can be solved despite their complexity, that costs can be recovered In large measure, and that such operations are socially and politically worth pursuing. 4. This report first describes the economic setting which has led to a re-examination of the contribution of cities and towns to economic development. From this analysis, a medium-term growth scenario is derived. The medium-term outlook places amphais on rural-urban exchanges, and on establishment in cities of a proper enabling environment for private Initiative. In this growth-oriented urban strategy, both rural and urban areas would respond more quickly to signals from domestic as vell as from external markets. Finally, specific measures are recommended to Improve revenue generation and financtal management at the local level, the management of urban land, housing finance, and urban transport so that resources in the sector can be spent more effectively. Actions are also proposed to strengthen sector institutions, so that the Improvements realized cam be sustained over the longer term. - iI - A. The Economic Setting of Urbanization S. In all developing economies, long-term economic growth at&d urbanization are interrelated processes. Sustained growth fosters urbanization and the division of labor between urban and rural areas. Conversely, a prolonged recession restrains internal migration and may eventually shrink the size of the domestic market relative to the subsistence economy. 6. The experience of Madagascar from the late 1970s until very recently has been one of prolorned economic recession, high unemployment, and low and diminishing household incomes. A Social Accounting Matrix (SAM) provides a useful portrait of this situation, for both rural and urban areas, as of 1988. The analysis is undertaken In the context of what the report calls the real economy, which encompasses all activities and transactions whether monetized or noia-monetized, recorded or unrecorded, formal or informal. A SAM shows how the incomings, or resources, of each economic agent or account are balanced by the corresponding outgoings, or expenditures. 7. From this analysis, the rural economy in 1988 is one in which, for the average farmer, the main source of money income derived from farming is his production for the domestic market, beyond his own food needs. The growth of urban areas, which account for most of the domestic market, is a necessary (though not sufficient) condition for the growth of agricultural productivity and rural incomes. In the urban economy, about half of construction, housing, and other activities takes place in the informal sector. .The informal sector has a vital role to play in filling the gaps and exploiting the niches left by the modern sector. For this to occur, however, the modern sector itself should expand, in ;rder to provide infrastructure, markets, and other parts of the enabling environment within which the informal sector can develop. Goods and services exchanged between urban and rural a -as are of about the same order of magnitude as goods exported, attesting further to the importance of the domestic market. 8. In this context, the total urban population increased from 0.7 million in 1960 to 1.8 million by 1984. This growth was at least 6% per year in most cities in the 1960s, but with the economic recession slowed considerably afterward. Antananarivo grew at 6.72 in the 1960s and early 1970s but at only about 4% by 1988. For provincial (Faritany) capitals and smaller townt the rate was slightly higher. When the growth in Antananarivo slowed, expansion of its urbanized area slowed even further, leading to higher densities and overcroNwding. One result was the growth of squatter settlements, in both Antananarivo and some Faritany capitals, as one of the few alternatives to overcrowding. 9. Local authorities, public enterprises, and the central government all struggled during the 1980s to bring direction and management to urban growth. However, few tangible results were achieved. Public expenditures for urban development by all levels of government have been extremely small since 1980, except those made under the first Urban - IIi _ Development Preject (PDUM) launched In 1985 with IDA assistance. Neither local governments nor the National Economic Development Fund (FNDE) has devoted much resources to urban development. Under these conditions, urban population growth has had to be accommodated through increased densities in existing neighborhoods, and to a small extent by private construction on penr-urban egricultu:al land near road and utility networks. 10. Viewed from a wider perspective, however, the 1980. were less of a disaster for ur-an policy. Important lessons were learned, both at the policy level and from the practical experience of operations. Despite an unfavorable economic environment and some false starts, the 1980s saw the introduction of several key elements of a consistent urban policy. The main challenge for the 1990s is to develop these elements further and combine them in a more effective partnership between the central gover: .Jent, local author'ties, and the private sector. B. A Growth-Oriented Urban Strategy 11. Resumption of sustainable economic growth, with a manageable balance of payments and relative price stability, is the main strat'egic objective of the Malagasy Government. The program of stabilization and macro-economic adjustment implemented by the Government with the support of the World Bank and the IMF is expected to help achieve positive per capita income growth of the order of 2S per year over the long term. Up to the present, the adjustment program has emphasized the need to promote the growth and diversification of exports. In addition to external trade, however, a revival of domestic trade is critically important for sustained economic growth. Domestic trade is best promoted through the intensification of rural-urban exchanges, and by investments in the necessary Infrastructure and services. 12. The elements of the SAM analysis for 1988 were projected to 1998 under assumptions of low and high urbanization. The high urbanization scenario, believed to be the more likely outcome, is associated with a revival of both foreign and domestic trade. Per capita agricultural and rural incomes, not just those in urban areas, are higher in the high urbanization scenario. Rural incomes grow at about 1.42 per year, compared with 0.22 in the low urbanization scenario. The difference is wholly accounted for by the domestic market, since both scenarios assume the same growth of export crops. Rural areas that are connected to growing urban markets can increase their commercial production, which generates cash income, new patterns of expenditure, and a market for related activities. 13. In interpreting these results, it should be remembered that the SAM model does not provide an explanation of growth, but instead gives a view of economic interactions at a given time. A full model of economic growth would include these relationships, but also others such as relative prices and production coefficients in the different sectors of the economy. Calculation of a SAM, especially at two points in time, can however be useful in providing Insight into the behavior of economic agents. This - IV - behavior underlies the model and explains the difference between the 1988 base situation and that pro.pectod for 1998. 14. In the model, between these two time periods people continue to come to cities, In search of better opportunities and an improvement In their standard of living. In the city, they are not the same economic agents as they were before. They are no longer able to produce food for self-consumption, so must purchase It on the market. They need a higher income to sustain the necessary increases in expenditure. At least initially, migrants would not have this Income and would therefore draw on the family's earlier savings or from transfers from other urban dwellers and from the village (remittances). Nowever, there Is no evidence of a continuous Increase in inter-household transfers over time, so it is likely that this call for transfers Is transitory. Over time, income Is earned in the city. from a variety of production activitiess street vending, repair services, caretaking, and the like. Such activities could not successfully be launchod In rural areas. Pushed by the need for cash, the newcomers take advantage of opportunities, make use of new Inputs, and benefit from economies of scale, mainly because they have no alternative. 25. However, there is nothing automatic about these changes. The results generated by the SAM model are only possible if the physical and economic environment, created by past Investment, is favorable. If this Is not the case, few people would take the risk of moving to cities which offer no opportunities. Nor are the changes without cost. New migrants increase the pressure on the existing stock of Infrastructure and services. If private investment stagnates, and If the expenditures of central and local governments do not inerease, everyone becomes worse off through overcrowding and congestion. 16. Public expenditures in the urban sector need to increase, in order to acconmodate migration In ctties which at the same time make the maximut coatribution to growta. In this regard it is useful to separate the expenditures that primarily benefit the city from those that primarily benefit the nation as a whole and are carried out in the city. Expenditures primarily benefiting the city are termed Local-Purpose Investments (Investissements i fonction locale, InL). 17. From the analysis of the current public expendituro program, it Le clear that the pace of Identification and preparation of new projefts, and of effective Implementation, is much less than is required in the sector. Opportunities for growth continue to be missed. More effective prograSuing of public expenditure would consist first of preparation of a long-term indicative plan, that would provide a vision of the future to which Malagasy societ*y is tending. This effort could be incorporated Into a national strategy for implementation of recommendations of the Long-Term Perspectives Study (LTPS) particularly applicable to Madagascar. Medium- torm objectiver supporting this future vision could be contained in a devolopment plan, that wvuld translate and adapt the long-term vision to the present macro-economic environment and to existing constraints and opportunities. The development plan would be indicative, largely for the purpose of urban and related planning, and focused on geographic and spatial distribution, and should not be confused with centralized planning of the entire economy. Within this framework, the review of the three-year Public m v - Expenditure Program (PEP) would be conducted to verify that the cost of the program can be met with the resources avai' 'le, and to monitor their lmplementation. 18. rFr. expenditures in urban areas would be devoted both to rehabilitat'^n of existing infrastructure and services and to accommodation of new mas.. tc Based on plausible assumptions about IFL expenditures, it ts conclu4 . that about one-third of asuh expenditures should be devoted to expansion oz 4.ban areas, whils fully two-thirds would be required for renovation c' existing facilities. C. Strateatv Imnlementation 19. For urban expenditure programs to be successfully carried out, sound strategies and strengthened institutions are required in the most Important sub-sectors of urban development. Approaches and recommendations are proposed in the report to achieve this progress in local finance, urban land management, housing finance, and urban transport, as well as in the strengthening of urban institutions. These conclusions are briefly described below. Many of the specific recommendations are summarized in an Immediate Action Program, which proposes the short-term operational measures to be taken in support of the recomendations. Strenathenlna Urban Institutions 20. In order to promote an enabling environment for urban growth, central and local authorities will each need to be strengthened. At the central level, creation In 1988 of the National Technical Committee for Urban Development (CTNDU) has filled a critical gap in the institutional framework for urban development. It should be possible for the CTNDU to expand its consultative role, without interfering with the prerogatives of its member ministries. As supervising ministry for urban development, the Public Works Ministry (MTP) is responsible for urban policies applied by all central and local authorities. However, the internal organization and operations of MTP do not yet reflect the importance of urban issues. The Di&ectorate of Architecture. Urban Planning. and Housint (DAUR), in a restructured MIP, should be given important responsibilities that it would discharge upstream of program Implementation: identification of strategies; collection and analysis of urban datat assistance in management to local governments; urban planning provisions and guidelines; promotion of the domestic construction industry; and, at a later stage, post-appraisal of operations. A Directorate of Urban Operations (DUO), assuming the functions performed up to the present by the Urban Project Unit (BPU), would promote more efficient management and coordination of operations In the sector. Although its organization is satisfactory for the present, the Ministry of Interlor (MI) should similarly be strengthened in order to ensure that local authorities are properly organised, equipped, and staffed to make an effective contribution to urban development. - vI - 21. The various levels of local povernment (Collectivites decentralisies, CDs) have wide responsibilities, which, however, have not been clarified by defining the specific roles and powers of the different echelons of government within a given area. A reassignment of responsibilities is therefore an urgent priority, mainly to enable the mandate of city administrations to correspond more closely to the functional importance of the city. All towns of a given category would have the same responsibilities and resources, whether their governing CD is a Fivondronana ("municipality") or a Firaissna ("urban district"). There are also many cases where the city is spreading over two or more jurisdictions, and a single authority needs to be established. 22. This need for a single urban administration is most evident in Antananarivo. A two-phase approach is recommended to determine the most suitable legal and administrative framework for the capital. The first phase would focus on improved methods of municipal management, and on the creAtion of new area-wide structures to deal with specific problems such as flood control and urban transport. In this way, the city will gain experience with new methods of urban management on a regional scale, which it so far has lacked. In the second phase, the actual legislation would be considered and, if approved, the application of the new structure would begin. Local Government Finance and Financial Management 23. Urban management in cities and towns of Madagascar is first and foremost a financial issue, revenues being totally inadequate to cover even the most basic needs. Resources of local governments consist of central government transfers, local taxes, and other revenues. For central government transfers, the key principle should be transparency. The Ministry of Finance should commit itself to transferring the amounts to which the CDs are entitled, publish and adhere to a schedule for making the transfers, and consider a new allocation system based on fiscal performance and infrastructure needs. To increase the yield from local taxes, the tax base should be enlarged to reach people who presently escape taxation, and the effectiveness and efficiency of tax collection improved. Other revenues controlled directly by CDs, including revenues from goods and services and parafiscal revenues, have a good potential even though CDs do not rely heavily on them today. 24. Efforts devoted to securing more resources for local governments will have been wasted unless financial management at the local level is significantly Improved. Existing legislation and guidelines should be assembled in a "Code of Local Governments," and "Financial Management Guide," made available to CDs by the central government. Publication of a yearly note giving directions and targets is one of many relatively simple measures that would improve the quality of budgetary procedures. Improved expenditure programming and control will require that CDs make more realistic expenditure forecasts, meet all recurrent costs of operations, and, in a later phase, devote resources to investment using the Fund for Infrastructure and Urban Housing (FIHU), banks, and other external financing. - vil - 25. An expansion of trainint is required for local governments to become more self-reliant. More generally, training which emphasizes the transfer of knowledge, by means of manuals of procedures guidelines, and information, should be a critical element of any urban development program. Training by the Ministry of Interior for elected officials and central staff seconded to cities and towns should be strengthened. The Public Works Ministry should continue to organize workshops on urban development Issues, aad or. urban finance and management. In this regard, courses in municipal management should be added to the programs of the new public administration training institution (ENFAR). For each community, specific training programs based on an assessment of training needs should bo prepared following the Identification of investments to be undertaken. Urban Land Manaaement 26. Even though the rate of urban population growth is moderate, cities and towns have been unable to provide buildable sites in the numbers required. One Important reason is that only a small fraction of plots have fully legal tenure. In the vast majority of cities over the last 40 years, people have built houses on land that is either of uncertain ownership or registered to someone else. Subdivision controls and building permits are rarely applied. Moreover, families may feel they may occupy virtually any land without prior authorization, especially private State land which is highly vulnerable to squatting. 27. Future actions in urban land management should be taken in support of three main objectives. First, centralization should be reduced and treater involvement of local governments -romoted in the administration and management of both private and State lands. Seconds intermediate procedures toward the legalization of traditional land practices should be identified. The third objective is to introduce Incentives to facilitato the adogtion of these intermediate 2rocedures by both land occupants and local administrators. Reform measures should be introduced in three phases, of 4-5 years each. Actions in the first phase would Include a general cleaning of existing texts, mainly to eliminate inconsistencies and duplication. Drafts of the Settlement Permit, zone plan, and other new intermediate land-use instruments would be prepared for discussion. In the second phase, the new system would be introduced in more cities and towns that show commitment and lnterest. In parallel, a considerable effort would be devoted to resolution of uncertainties over who precisely is the owner or occupier of land (especially private State land) subject to squatter occupation. In the case of Antananarivo, responsibility for land registry, settlement permits, and urban plans would be transferred to the city administration. The third phase would consist of further transfer of responsibilitiy for land and urban planning to Fivondronana or Firaisana of major urban centers, with guidance from the Property Directorate (Direction du patrtmoine, DP) and other central agencies. - vill - Ron.in and Bous rt Finance 28. As part of efforts to revive the domestic market, there are good reasons to encourage investment in housing, both to make a positive contribution to overall growth and to alleviate the social impacts of the liberalization program. However, financial programs based on Improved mobilization of private savings are essential in order for urban housing to escape its current stagnation. Several programs of this type have been prepared over the past decade, b-it have not been carried out. Realistically, one should think not in terms of a single financing system but of a number of systems. In the near term, the quality of service and the ability to expand steadily while maintaining financial balance, rather than the number of dwellings financed, should be the goals of housing finance programs. 29. The most important element of the housing finance strategy is to increase the production of serviced lots, since this Is likely to be effective in increasing the number of savers and builders. The encouraaement of rental housina should consist first of allowing rent increases to follow the market, and next of granting temporary tax exemptions for new rentals. A policy favoring home ownership would have two main features. It would first assist in the establishment and operation of mutual or cooperative societies and associations for savings and credit. Second, con4ractual savings programs would be introduced on a competitive basis among institutions. Central Bank regulations should establish the guidelines, but not the details, for banks and other financial institutions to launch effective housing finance programs. Urban Transport 30. Low fares, poor roads, congested streets, and inadequate service have considerably reduced the supply of transport, especially in Antananarivo. A six-part program for the recovery of public passenger transport includes action to address issues of organization and management, fares, transport operators, and users. First, the management audits recently completed for the transport companies FIBATA and FDIM should be disseminated and their results subject to a detailed review. In addition, the idea of transwort plans is sound, but should be applied flexibly. As the capacity to prepare such plans is built up, scarce resources could be focused on the preparation of investment projects and their evaluation in the annual reviews of the Public Expenditure Program (PEP). Fares must also be raised, but discretion can be exercised over the speed at which tariffs covering recurrent expenses are reached, and measures taken to mitigate the social impacts. Fourth, the highest-priority improvements to the road and street network of Antananarivo have been identified, and should be considered in preparing PEPs. Fifth, studies and project planning on the scale of the Antananarivo metropolitan area shou'd be expanded, with the Traffic and Transport Bureau (Bureau de circula-on et de transport, BCT) likely to play a major role. Finally, to further encourage private initiativet, the 1988-89 legislation setting out the revised organizational framework for transport, and which allows independent operators as well as - ix m cooperatives and tiansport companies to be granted routes, should be given wid, practical applilation as rapidly as possible. D. IXqEpdlata Action Program 31. The actions outlined above constituto a medium-term program dsigned to revitalize Malagasy cities and the contributions they should make to economic development. Within this program are measures which should logically precede others, which facilitate reforms In related areas, or which for other reasons have highest priority and should be carried out first. These measures make up an I_ediate Action Program, and are summarized below according to the level of decision-making responsibility for them. At the national level: a. Reassign responsibilities among CDs and reclassify cit.ias and towns according to their economic lmportance (para. 5.6). b. Clarify and strengthen procedures for central government transfers to CDns and of the role of public accountantsl disseminate a Code of Local Governments, Financial Management Guide, and note on budgetary procedures. Impl_emnt a comprehensive program of strengthoning local finances in collaboration with selected CDs (paras. 6.10, 6.31, 6.44, 6.46). C. Implement first phase of urban land management program (pars". 7.28- 7.30). d. Launch process of preparation of long-term Image of Malagasy economy, as a first step in the process of preparation of a national development plan with PEP for urban areas (paras. 4.10-4.27). e. Take final measures to settle question of compensation to Antananarivo for "ssets of municipal corporation (EEN) (pars. 2.13). f. Establish mutual or cooperative savings associations and contractual savings programs, vith appropriate Central Bank regulations (para. 8.50-8.52). g. Reorganize NTP to give more focus to assistance from DAUR to CDs, and establish a Directorate of Urban Operations (DUO), assuming the tasks previously assigned to BPU, to coordinate and monitor project Implementation (para. 5.20-5.25). h. Carry out assessments of training requirements for CDs, national ministries, and specialized agencies, and design training programs in municipal management (paras. 5.33-5.311. In Antananarivo: i. Carry out the first phase of effort to agree on a new legal and admnlistrative status for the capital, including Improved management methods particularly by the Fivondronans of Antananarlvo-Renivohitra and creation of area-wide entities for flood control, transport, and other specific concerns (paras. 5.14-5.16). J. Establish a Technical Study and Coordination Unit for the Antananarivo metropolitan area (pars. 5.16). k. Following agreement with JIRAMA on amounts to be billed and other outstnding issues, achieve full cost recovery for water and electricity services (pare. 6.55). 1. Increase revenues from goods and services and market stall fees, and streamline parafiscal revenues (para. 6.36-6.41). In Faritanv capitals ind smaller townes m. Multiply urban operations of the Toamasina type, combining physical environmental Improvements and new housing sites with reforms in local finance and financial management, and the introduction of intermediate land tenure en urban planning instruments (para. 2.22 and 8.48). n. Conclude Program-Contracts with central government defining yearly objectives and the means of achieving them (para. 6.68). O. Increase revenues from goods and services and market-stall fees, and streamline parafiscal revenues (pare. 6.36-6.41). p. Followtng agreement with JIRAMA on amounts to be billed and other outstanding issues, achieve full cost recovery for water and electricity services (para. 6.55). PB"I. THE ECONOMIC SEIMN5 OF URBANIZATION CHAPTER I THE REAL ECONOMY OF MADAGASC' 4 AN VRBAN-RURAL PERSPECTIVE A. Introduction 1.1 In all developing economies, long-term economic growth and urbanization are Interrelated processes. An uninterrupted period of economic growth fosters urbanization and the division of labor between urban and rural areas. Conversely, a prolonged recession restrains internal migration and may eventually reduce the size of domestic markets relative to the subsistence economy. 1.2 Since the late 1970s, Madagascar has experienced a prolonged economic recession, as shown by per capita GDP which has shrunk by 50X in ten years. During this long process of economic decline, the modern sector, which is mostly urban,. and the urban economy as a whole have been strongly affected. High unemployment, low and diminishing household incomes, and Inadequate public services have considerably reduced the level of activity of enterprises serving the domestic market. 1.3 This economic decline affected urban areas in two important respects. First, it induced a significant abatement in the pace of urbanization and of inter-regional migration. For example, recent surveys of the Antananarivo metropolitan area 11 show that the 1988 growth rate of Antananarivo was slightly below 42, which, by the standards of African countries, is low. Smaller towns live Tsiroanomanididy, located outside the main centers of activity, have growth rates which are probably below natural population growth, turning them into centers of outmigration. Coastal cities and towns, however, seem to have maintained higher growth rates, in the range of 5 to 6%. Because of their connections with the rest of the world, they may have been less severely hurt by the crisis. Overall, the 1980-88 urban population growth rate is estimated at only 4S, compared with 62 during the 1960s and 1970s. 1.4 The second major effect is in the form of missed opportunities. Urban-based activities are responsible for 45% of the GDP. This apparent reduction in the pace of population redistribution can be seen as one form of adjustment to the economic recession. One might believe that low urbanization would help the economy by reducing urban poverty and unemployment and allowing local governments to improve their financial situation. In reality, however, an abatement in the process of population I/ In addition to the Fivondronana of Antananarivo-Renivohitra, the Fivondronana of Ambohidratrimo, Atsimondrano, Avaradrano, and a portion of Arivonimamo are included in most statistical and policy definitions of the Antananarivo metropolitan area. 2- redistribution between regions, and between rural and urban areas of Madagascar, represents opportunities missed to take advantage of the changing economic potential of different regions. Better management of the domestic market would have allowed some of these benefits to be realized. 1.5 Liberalization has placed Nadagascar on the threshold of a new phase of growth. If this growth is sustained through _he 1990s, urban areas will be called upon to generate as much as S5% of the overall GDP and to be responsible for 702 of the total GDP izs4rease during the decade. This would also imply that fully four-fifths of total capital accumulation would occur in cities and towns. The implementation of adjustment programs suggests that this formidable task would be carried out in an increasingly sound macroeconomic environment. However, the 1990s will also be a decade of severe financial constraints and difficult choices to be made among pressing needs in all sectors. It is therefore critically important to have a clear understanding of the present situation of the urban economy, of its present and potential future role, and of the costs and benefits of urbanization. 1.6 The purpose of this chapter is to analyze the contribution of Malagasy cities and towns to the overall economy. The analysis is undertaken in the context of what this report calls the real economy. The real economy encompasses all activities and transactions, non-monetized as well as monetized, unrecorded as well as recorded, informal as well as formal. Therefore, the aggregates of production, income, and expenditure generated by this type of analysis will be broader than the usual macro- aggregates compiled in national accounts. This also means that such figures, foL the time being, are not immediately accessible from existing information systems and require special analysis. The extra effort, however, is justified since only by taking account of all economic transactions can supply and demand in the domestic market be reliably estimated. 1.7 Section B below briefly presents the overall situation of the economy in the base year 1988. Sections C and D are devoted to the rural and urban portions of the national economy, respectively, for the base year. In Section D especially, attention is called to the critical weaknesses of the current urban economy, such as the low productivity of urban-based activities and the inefficiency of the transfers that cities receive from the rest of the country and from the rest of the world. The concluding section of this chapter reviews the components of rural-urban exchanges and explores the relationships among population redistribution, urban-rural exchanges, and economic development. B. National Resources and Exnsnditures in 1988 1.8 As a benchmark for the analysis, a matrix of population distribution (shown in Table 1.1 below) shows the urban vs. rural distribution of the population according to socio-economic categories, togethe- with the corresponding growth rates. These particular socio- economic categories are adopted because they show important differences In consumption and investment behavior, and therefore would react,differently to policy changes affecting the urban sector. The primary population is mostly engaged in agriculture. The mixed population, although also engaged in agriculture, hAs other productive activities such as trade. The informal non-agricultural population is the difference, in both urban and rural areas, between those cl"ssified in other categories and the total population. Employees of the central ard local governments (except the top levels, included in the elite, and employees of public enterprises, in the 'other modern v :tor" category) msas up the civil servants category. In the other modern sector are those deriving most of their income from (non- agricultural) wages and profits or ownership of firms or enterprises. Finally, the upper level of ths society is identified as the elite, which plays a major role in overall expenditure patterns and in the decision- making process. 2/ As discussed further in Annex 1.1, the level and pattern of expenditure of each category of the urban and rural population is quite different, especially regarding their sensitivity to changes in the macroeconomic environment. 1.9 Table 1.1 indicates that the urban population, defined as the population of the 100 cities and towns with more than 5,000 inhabitants in 1988, represents 26Z of the total population. The urban modern sector broadly defined, including civil servants and the elite, comprises 362 of the urban population and its present growth rate is of the order of 32. at 2/ Expatriates from developed countries are considered as part of the elite category. at This growth was probably negative through the 1980s, during the worst of the economic recession, until about 1987. - 4 - Table 1.1 DISTRISTION OF TOTAL POPULATION IN 1988 ('tw URBAN RURAL TOT.POP. GROWTH RATES 88-" X POPUL. I POPUL. s POPUL. U R P AGRICULTURAL 7 2W 68 4,601.6 46 6,862 2.0 2.? 2.7 mIXED 10 460 80 8,000 81 8,450 4.0 8.0 8.1 INFORMAL 42 1,200 4 s0e 18 1,600 6.2 2.0 4.0 CIVIL SERVANTS 11 8a0 2 140 4 470 8.0 2.0 2.7 OTHER MODERN SECTOR 22 041 1 so 0 701 8.0 2.0 2.0 ELITE 2 680 a 4.2 1 57.2 2.6 2.0 2.0 TOTAL MODERN 80 1,024 2 204.2 11 1,226 TOT.NON AGRICULTURAL 77 2,224 0 604.2 24 2,726 TOTAL 100 2,674 1in 6,3BO 189 11,240 4.0 2.8 8.1 X TOTAL POPULATION 25.6 74.4 108.0 Source: Total population: World Bank projections based on 1985 Census. Urban population: Mission estimates. 1.10 GDP, calculated to include non-monetized as well as monetized and recorded as well as unrecorded transactions, may be termed 'real GDP.' Table 1.2 shows the sectoral distribution of "real GDP.' The estimates of the table use the broader concept of the real economy, and therefore differ (often considerably) from official national accounts. For example, Madagascar's 1988 GDP at market prices as conventionally estimated sounted to FMG 2,650 billion 41, while the SAM estimate Is 162 higher. The share of- construction and public works in GDP is 1.8% as assessed by the Bancue des donnees de l'Etat (BDE), but 6.62 when, all transactions in the real economy are accounted for. 1.11 The larger estimates given by the SAM for the value added of some sectors and for GDP as a whole do not, however, support the conclusion that the Malagasy economy is in relatively good shape. The larger per capita GDP estimate for Madagascar of US$196 (see Table 1.2) can be compared with estimates for Zaire and Cate d'Ivoire, using the same real economy approach, of US$359 and $1,165 respectively. Moreover, the estimate for the share of construction and public works, although five times higher than the BDE estimate and 2.5 times the World Bank proje tion for 1988 in the revised national accounts, is less than one-sixth the corresponding figure for Cote d'Ivoire, which has roughly the same population. The over-valuation of the CFA Franc undoubtedly explains part of this difference. The rest, however, reflects the relative weakness of Madagascar's construction sector, which is a good indicator of the real welfare of households and a strong influence on their productivity. The low levels of private and public investment are illustrated f- Antananarivo, where more than two-thirds of the population 41 World Tables, 1989-1990 edition. increase is accommodated in existing housing units. The floor space per inhabitant in Antananarivo is 3 m2, very low by African standards. Table 1.2 GRDSSOMC PROC BY SECTO (IN US$ MILLION) YEAR 16 NON MONIYZE MONETIZED TOTAL RA GOP MARKET PRICES FCT.COSTIIT PRICES FOOD SELF CONWTION 414 410 416 OTR AGRICULTU 448 401 448 MINING METALJRGY 14 8 14 MAIUFACTURIN 214 192 214 CNSTRUTION PUSLIC WORKS 48 16 186 140 UTILTIT so 27 80 TRANSPORT AN TRADE 859 $10 8C9 PRIVATE SERVICES 186 281 $82 see ADMINISTRATION 116 115 116 IMPORT TAXES IS66 GROSS DOMST.PRODUCT US MILLION 691 161 169C 2208 GDP PER CAPITA USS PER INHABITANT 68 144 174 196 ROSS DOMST.PRODUCT mG MILLION 844191 23094 2791066 81S226S Source: Mission estimates 1.12 Table 1.3 shows the relative contributions of urban and rural areas to sectoral value added, and to GDP. Average value added per capita is US$342 in urban areas and US$146 In rural areas. The comparable figures for Cote d'Ivoire are US$1,717 and US$723 res lctively, or five times higher In both cases. Interestingly, the urban-rural disparities are the same In both csses. If the analysis were taken further, and comparisons made at different points in time, the impact of development strategies on urban- rural income differences could be traced. A premise In this respect suggested by the present analysis is that the actions undertaken In agriculture to promote the development of Madagascar have had little impact on the relative productivity of rural and urban areas. 5/ 51 A corresponding premise c,n be advanced for Cote d'Ivoire, that the "urban bias,, allegedly very strong, has not in fact Induced a larger urban-rural difference in productivity than is found In Madagascar. - 6 - Table 1.3 URBAN AND RURAL CONTRIBUTIONS TO CDP (IN US$ MILLION) YEAR 1988 ORANC# CONTRIBUTION TO VALUE ADDED URBAN RURAL TOTAL URB/TOTAL FOOD SELF CONSUWMPTON 57 859 416 0.187 OTHER AGRICULTURE 79 864 448 0.179 MINING METALLURGY 12 2 14 0.806 MANUFACTURINO 121 9J 214 0.667 CONSTRUCTION PUBLIC WORKS 76 71 146 0.614 UTILITIES 28 2 80 0.930 TRANSPORT AND TRADE 220 181 869 0.685 PRIVATE SERVICES 222 144 806 0.666 ADMINISTRATION 86 20 116 0.764 IMPORT TAXES 76 28 105 0.729 TOTAL REGIONAL PRODUCT 984 1222 2208 0.448 VALUE ADDED PER CAPITA 842 148 198 U/R : 2.84 Source Mission estimates. 1.13 The pattern of expenditure of urban and rural households, although rather similar to that observed in most countries of Sub-Saharan Africa (SSA), reveals that the standard of living of the modern sector is depressed (see Tables 1.4 and 1.5). This can be seun by comparing per capita consumption (including non-monetized consumption) for the different population groups, and locations. Taking this figure for the more remote rural areas as an index of 100 (corresponding in 1988 to US$90 per capita), other averages range from 130 for the agricultural population as a whole, to 200 for the urban informal group, 460 for civil servants and the other modern sector, and 3,500 and above for the elite. As Table 1.5 shows, urban areas, which account for 47% of total household expenditure, are responsible for fully 852 of total imports for final consumption. -7- Table 1.4 TOTAL HOUSEHOLD EXPENDITURE IN 1988 (IN US$ PER INH BITANT AND USS MILLION) URBAN RURAL TOT. POPULATION PER INN. TOTAL PER INN. TOTAL PER INN. TOTAL AGRICULTURAL 128 26 117 570 118 S9$ MIXED 174 78 1" 402 189 481 INFORAL 188 220 1in 40 174 261 CIVIL SERVANTS 480 142 3a4 48 a9s 185 OTHEIR OERN SECTR 449 284 851 21 486 W8 ELMTE 8848 264 188T 0 a88 210 TOTAL MODERN 01s 086 840 09 669 09 TOT.NQ4 AGRICULUML 882 880 218 110 852 800 TOTAL. POPULATION 882 054 120 1082 181 2to X OF TOTAL WEND. 47 68 160 Sources Mission setimtes Table 1.5 HOUSOLD EXPENDITURE ON IMPORTED GOODS IN 1988 URBAN RURAL TOT. POPUTION PER INN. TOTAL PER INN. TOTAL PER INH. TOTAL AGRICULTURAL 8 2 2 11 J is MIXED 14 0 4 11 6 10 INFORMAL 19 22 7 2 10 24 CIVIL SERVANTS 78 24 29 4 eo 28 OTHER MODERN SECTOR 91 58 49 S 87 S1 ELITE 1887 1 1267 72 TOTAL MODERN 16 158 42 8 132 162 TOT.NON AGRICULTURAL 79 176 21 11 08 186 TOTAL POPULATION 04 188 4 88 19 210 SHARE OF IMPORT IN EXPEND. 22 S 15 X OF HOUSEHOLD IMPORTS of 1S 160 Sources Mission estfLates - 8 - Social Accountint Matrix (SAM) of the Nationdi Economg 1.14 With an urban-rural breakdown of the population, and a suitable accounting framework, It is possible to show the Incomings and outgoings of resources both for the nation and for the population that is specifically urban. Such an accounting framework is provided by the Soctse. Accounting Matrix (SAM). Table 1.6 presents an aggregated version of the SAM for 1988 covering the whole country. In essence, a SAM shows how ths incomings, or resources, of each agent or account are balanced by the corresponding outgoings, or expenditures. For a fuller description of the mechanics of this approach, see Annex 1.1. 1.15 The results of the SAM will now be examined in more detail so that implications can be drawn for the behavior of the rural economy, the urban economy, and urban-rural exchanges. A profile of the economy of AntAnAnarivo, and of its exchanges with the surrounding region, is presented In the concluding section. C. The Rural Economy and the Agricultural Sector 1.16 In rural areas, agriculture is the mainstay of 94X of the population. It generates 672 of total household income, and 902 of the private expenditurea are made by farmers. These figures demonstrate that agriculture Is the driving force of the rural ecorzmy. They do not, however, indicate the locations of markets for agricultural production. This question is addressed below as rural productivity, income, and growth ere discussed in the framework of the real economy. 1.17 Table 1.6 sbows that, at farm gate prices, export crops (including export taxes) at US$174 million represent about 172 of the total agricultural output of US$1.03 billion, a proportion comparable to those found ln other SSA countries. 6/ Par more important is the production of food and other primary commodities, such as fuelwood, for the di.mestic market. However, it is critically important to note that nearly half of the total production for the domestic market is not monetized. It is consumed locally, with little or no effect on other sectors or on the monetized economy. For that portion which Is monetized, the domestic market accounts for 672, and export crops the remaining 332. The domestic market is shared roughly equally between rural areas (452) and the cities and towns (552). 61 Defined from Table 1.6 as agricultural commodities which are exchanged with the exterior (283) divided by total agricultural commodities (1682). TADLe 1.6 uci A**nlM NUtM O TM ,W.Dt GiN7 IF 1FE WMI*N 666 MIN.. Aifl FRIC I 2 a 4 6 9 10 1it 1 is 14 is to 6 to9 t OASIoS W oSleTS vman 6046176 A?vtT CDm.L tClUWE MAL octet mzm tWlS NA*L LoC.O IItl.0WTD &CO.OH "M. USIC. -l A.6lu1m ST 011r06c 16MA. ".C.<6Dfc9.aw Cum NW aW CWITAL TTAL A I_l Hn S go 2 2 o 184 72 66 a# -0 9S a SAL I_usLIS 6 6 a a 6" as 34 low 4 am C t- 0_6S 1is I 6 4 7 0 0 1 6? I a 0 6066t. 406 as 2 li 12 18 * U gm 1? so E _QS1 FO s. OM.m * 511t 2 p vau c .m"on6 si 191 go 126 - 62 a WAMIS D _1366 1a 6 115 so 0 H W.IC. 120 to a U 29 140 2 low t 6.Om 6I Ua1n1 ea 497 1i to 176 J AC.Mv.uIaA. io o 16 am K A11v.0IS1UIcTm 0 - - L IV.01 I1 a? 16 N CW.A=.HJWN U 2 45 SW 0 17 N CW.L.LCA QM# 6 0 a 6 0 CAP.AC.W4ul 65? am U 7? l15 P TMAL C0UR< 9w 106E 8 200 66 as 1 10 106 28 1a" 1V 6 I1 R 9TASR.CI 1N 0 a4 so 41 * 0C!60.CAML O tt 7 T 0TL 964 I0 U so J 2 1260 6n N09 166 1 16 5 2 16? 6 i1t 477 7 - 10 - 1.18 For the average farmer, therefore, the main source of money income derived from farming is his production for the domestic market, beyond his own food needs. In constant prices, the growth rate of the urban market for food and other agricultural products is of the same order of magnitude as the growth rate of the urban population. With a high urban population growth rate, for example, the growth of the average farmer's income generated by the sale of food to urban consumers is likely to be greater than what can be expected from the sale of export crops. 1.19 Non-agricultural sources of Income in rural areas, which account for 332 of rural income, are also linked to the dynamics of the agricultural sector. Over time, non-agricultural income per person can grow only if agricultural income itself is growing. Artisan or small-enterprise activity for the rural market should be developed as one element of a rural strategy, but by itself would have limited effects because of weak purchasing power and markets for the products, and (in areas of low population density, with poor infrastructure and services) low productivity. The growth of urban areas supplements the rural markets with additional consumers, further enhancing agricultural productivity and rural incomes. 1.20 Many of these features of the rural-urban economy are illustrated in the Antananarivo region. The periphery of the capital city consists largely of plains of the Ikopa, Mambs, and Sisaony rivers and their tributaries. where irrigation has been established for more than a century. Although i. _gation implies a rather well-developed form of land tenure and agriculture, tl'e management of local governments in these areas has been poor. This has forced many residents to retreat from the market economy to a pattern of activities having more to do with immediate survival and local self-sufficiency. The rural periphery of Antananarivo is crowded with candidates for urban in-migration which the city is unable to accommodate. There is consequently the paradox of a low development of commercial agriculture, very small family farms, and low money incomes on the outskirts of a major city. D. The Urban Economp 1.21 At first glance, urban areas, which contribute 45X of total GDP, 50S of total savings, and 662 of total investment, appear in better shape than the rest of the Malagasy economy. However, a more detailed analysis by type of household and by activity reveals the precarity of the socioeconomic situation in cities and towns, especially those of the central plateau and the smaller towns. The rapid deterioration of infrastructure and public services, combined with inappropriate macroeconomic policies, have led to the decline of construction and public works, private services, household goods, and other urban-based activities directed mainly at the domestic market. Equally as important has been the resulting contraction in urban- rural exchanges. - 11 - 1.22 The disparity of per capita expenditure between urban and rural areas noted earlier (see para. 1.12) is the result of several complementary factors. The distribution of the urban population by social category differs significantly from that in rural areas. Cities also typically require higher expenditures for the satisfaction of basic needs, such as housing. There may also emerge a different perception in cities of the desirable level of consumption: by offering a larger variety of goods and services, the cities generally increase the willingness of residents to buy and, consequently, their willingness to earn cash income. Examples of these effects observed in urban areas are the larger share of services In the prices of consumer goods such as food, higher perceived needs in public infrastructure and services, a higher share of monetized consumption in total consumption, and a higher share of consumption imported from abroad and from other regions of the country. In Madagascar, however, these effects are only beginning to be felt. Incomes of most urban households are so modest that food expenditures still make up 59X of total expenditure in the modern sector, and reach 74Z in the informal sector. Little purchasing power is left for other goods and services which generally form the basis of the urban economy. 1.23 In cities and towns of Madagascar, three main sub-sectors of specialized activity explain a large share of total urban output as well as the extent of urban-rural exchanges. First, agroindustries, which are mostly devoted to feeding the cities, account for some 402 of urban output. Second, construction and housing, devoted to building the cities and sheltering inhabitants and their activities, makes up about 202 of the total urban output, even though this sub-sector i.s depressed and should be contributing much more to output. Finally, administration, devoted to governing the entire territory through the urban network, accounts for about 102 of urban output. 1.24 About half of these activities take place in the informal sector. Although cities and towns will rely heavily on the informal sector to become revitalized, its potential contribution should not be overestimated. At present, it accounts for only about 202 of urban value added. Over the long term, the informal sector has a vital role to play in filling the gaps and exploiting the niches left by the modern sector. For this to occur, however, the modern sector itself needs to expand, in order to provide infrastructure, markets, and other portions of the enabling environment within which the informal sector can develop. E. Urban-Rural Exchanges 1.25 The exchanges between urban and rural areas form one major part of the long-distance exchanges which take place inside a developing economy. These exchanges involve movements of population and labor, transactions of goods and services, and transfers of revenue and capital. - 12 - 1.26 Exchanaes of Population. As mentioned earlier, population redistribution, involving net migratory flows from rural to urban areas, is the main determineant of the division of labor and of urban-rural exchanges in SSA. Urban population growth, however, is one among many complex movements of populatior.. The same urban growth rate may correspond to different stages or paces of the division of labor. In Madagascar, the decline of the Image of cities as poles of modernity and their difficulty :n accommodating new migrants has promoted a sort of Iruralization" of the urban peripheries, in which commercial agriculture tends to be replaced by low-productivity farming for food self-sufficiency. 1.27 Exchanges of Goods._Services, and Income. Transactions between urban areas and rural areas of Madagascar appear in the urban SAW presented in Table 1.7 below. The upper left portion of this urban SMH shows the transactions taking place within and between urban areas themselves. The lower left part reflects the payments made by urban areas to rural areas and to the rest of the world. The upper right poraton summarizes the incomings (receipts) In urban areas from rural areas and from the rest of the world. Transactions of urban areas with rural areas and with the rest of the world are categorized in the second table of the urban SAM (Table 1.8). Table 1.7 SKIM. _cnI o K%nUO e wow _, IWLs ,arc iw ie "S=M~ ~mri. NO 1RRMSATflJ UOIg iMO Md.m PAW MMSN IE MM OF 1TM VWb waia Pe we meum. RN 0~n MMI weuws LOCIL mOM. FMr we.CMs.#oUeW. MIC. siwe ANUM..asMm suM ,uMw.. c.GMrWi.OW am. AMU M ma am "Wmui OMAL OM femaOS s * in so as we a n we OCLX _M.1 is 4 0 S O 0 A as * a so 0Km. fIN.W u 1, 54 4 4 05 M4 47 1J t t M M. Me 2 onWI WNW$ ma.m1M Ma as S' 4w marsIUU lit 6 at M0 0 c...iAWM. on as a a as as as lae Me W :a oWI Sto uo "MW a cc..MM 0 4 " U N 1"FIV.Co umww jl " JO #1 I t40 M 104 t COP.M.ISM. 10 5 1 0 U CnaC.Lc & imra as TWAP.. vapw a mese a e e w e u e aaas ee h aar urs To M.L MaS a lte lit M .3 ass ONw pAWOS To a.m OF110 UND is 40 4? we 8k as. ALOte CAIMA. IeGWe I n aso mPI MoAIL we se us8 am 429 we Mer u. Ws 1e to" lit a se e as" in - 14 - ji aRI8I!r I; l J S 5a 30 313e aE t. 000 £ 0 0 S. 4 Its in local finance (resource mobilization, tax collection, etc) and local financial manatement (budgetary procedures, expenditure programming and control, management Information systems, etc). Annex 6.1 presents a description of the different layers of local governments in Madagascar, with their structures and responsibilities. LOCAL GOVERNMENT FINANCE 6.3 The local finance system should enable the resources of local governments to exceed their current expenditures, thereby creating a self- financing capacity sufficient to carry on essential investments. During the last ten years, this minimum level of financial autonomy, and therefore of satisfaction of local needs, has never been achieved. The quality of the services provided by local governments has continually eroded, so that it often appears that the main function local governments perform today is to pay their employees' wages. The erosion of local revenues and the - 60 - irregularity of central government transfers have prevented any kind of planning and rational management. Therefore, resource mobilization is now a top priority item. Local governments should aim at tripling their resources (central government transfers excluded) during the next five vears. 6.4 All the analyses presented in this document focus on the "ex- Communes Urbaines" (Annex 5.1), since these urban entities represent more than 80 percent of the total urban population and since information on them is readily available and reasonably detailed and accurate. 6.5 There are three main types of resources for local governments in Madagascar: (i) Central Government transfers: these constitute the biggest share of local government revenues in Madagascar (more than 50 percent in 1988). They consist mainly of transfers of tax receipts to local governments: 18 percent of the Transactions Tax (Taxe Unique sur les Transactions, TUT) and 8 percent of the Internal Consumption and Import Tax (Annex 6.3). In addition to these taxes, the Central Government also gives outright subsidies to local governments from the general budget; (ii) local taxes, consisting of the land tax, the property tax and the business tax. These taxes are mainly managed by the Central Government, which transfers their proceeds back to the Fivondronana and Firaisana in the case of the land and property tax and to the Faritany for the business tax (Annex 6.3). The property tax is a major stable resource for the urban ?ivondronana/Firaisana and often represents about 20 percent of their revenues; and (iii) other revenues, comprising revenues from goods and services (rents from municipal buildings, cesspool draining fees, etc) and of parafiscal revenues, i.e., extra-legal charges and taxes (market stall fees, taxes on shows, alcoholic beverage sales, ete). They generally represent between 10 and 20 percent of local government revenues. Annexes 6.5 and 6.6 present a detailed breakdown of revenues and expenditures for the Fivondronana and Firaisana 'ex-communes urbaines" in 1987 and 1988. There are few differences in the structure of revenues and expenditures between the two types of entities. Size also seems to have little impact on the balance between the three types of resources defined here. Central government transfers, for example, varied between 58 and 73 percent of Fivondronana resources in 1988 (except for Toamasina with 41 percent). For the Firaisana, they represented between 63 and 78 percent of resources for the same year. 6.6 Most of the responsibility for managing urban taxation in Madagascar is in the hands of the Central Government, which generally determines the ta. base, maintains tax registers and collects taxes. This is particularly true for the categories of revenues (i) and (ii) above, which - 61 - are mainly influenced )j factors external to local governments. The revenues on which local governments have most control are those defined above as "other revenues", which are mainly dependent on internal factors. Each of these types of revenues raises specific issues, and are therefore analyzed separately below. A. Central Government Transfers 6.7 In developed as well as developing countries, expenditures of local authorities tend to grow faster than revenues. This financing gap is generally filled with transfers which are not tied to any specific expenditures. In Madagascar in 1988, these transfers represented more than 50 percent of revenues for urban Fivondronana and more than 60 percent for the Faritany and the urban Firaisana. Despite the size of these transfers and their political, social and economic implications, the criteria for allocating them have not always received the attention they deserve. The following analysis focuses on three issues: the volume of transfers, the periodicity of transfers and the criteria of allocation. 6.8 The volume of transfers. The Faritany, Fivondronana and Firaisana 'ex-communes urbaines" are supposed to receive a portion of certain tax proceeds (see pars. 6.5 [ii] aad Annex 6.1). In 1985 and 1986, the Faritany received respectively 14.2 and 13.2 percent of the TUT instead of 18 percent, representing a for.one income of FMG 1.7 and 2.4 billion. In 1988, the "ex-communes urbaines" should have received FMG 7.5 billion from the Internal.Consumption and Import. Tax; they in fact received only FMG 6.3 billion i.e., a foregone income of FMG 1.2 billion. 6.9 It appears that the Central Government does not always fulfill its obligations toward local governments. The volume of transfers seems to be determined by the amounts included in the appropriation law and not by actual tax revenues. Moreover, the level of these tax revenues is confidential information; elected officials are therefore not able to claim the additional revenue they are entitled to. That makes budget forecasting a difficult exercise for elected officials. Either they overestimate the transfers and wind up with a deficit at the end of the fiscal year, or they are conservative and underestimate the transfers; but then, legal constraints prevent them from spending the excess money they receive. There is also the issue of the claims accumulated by local authorities on the Central Gover=nent during the 1980s, since the transfers never reached the level stipulated by law. 6.10 The key principle for Central Government transfers should now be transparency. A number of steps should be taken: - the Ministry of Finance should commit itself to respect the texts and transfer to local governments the exact amounts they are entitled tol - 62 - - the Ministry of Finance should communicate every year to local governments' estimates of its tax receipts early enough for them to make sensible budget forecasts; - during the year, the actual receipts should be communicated to local governments on a quarterly basis, so that they can plan for the rest of the year; - the preparation of a compendium of legislation on local government receipts wav4I greatly help the CDs evaluate the possibilities for improve?d receipts; and - the issue of local governments' financial claims on Central Government should be addressed. The Central Government is now working along these lines. For instance, Central Government transfers to the Fivondronana of Antananarivo-Renivohitra hAve trebled, from FMG 700 million in 1986 to FMG 2,030 million in 1988. Such efforts should be encouraged. 6.11 The periodicity of transfers. There is no official schedule for Central Government transfers. For instance, In 1988, the first payment to the Fivondronana was made in April (two months later than the usual date). A second payment was made in July, a third in October and a fourth in December, i.e., too late to finance any expenditure for that year. Three of the four payments were made in the second semester, pushing all expenditures towards the end of the year. 6.12 Cash management at the Central Government level is a complex task. Nevertheless, it should be possible to better protect local governments' prerogatives by: - establishing and resoectinat an official schedule of transfers. One possibility would be to switch to monthly payments equivalent to one twelth of the budgeted amount; and - settling any outstanding transfers during the first quarter of every year, once the actual receipts from the TUT and the Internal Consumption and Import Tax are known. 6.13 The criteria of allocation. Today, there is a unique criterion of allocationt population 23/. Population is without doubt a very Important factor, but it is often unreliable and furthermore has no 231 Ordinance 73-015 of March 30, 1973, Article 5: "... La repartition des sommes revenant aux communes urbaines, en application des dispositions qui precedent, sera operee annuellement en fonction du dernier chiffre officiellement homologue de la population de chaque commune."N - 63 - incentive function for local governments. In effect, Central Government transfers could be a tool to improve local governments' effectiveness and efficiency, by linking the amount of transfers to local governments' performance using a number of criteria to be determined. A new allocation system could rely on two main criteria: fiscal performance and infrastructure needs. In this regard, it would be useful to review the distribution of receipts from the Transactions Tax (TUT), to see if the rates of transfer applying to certain cities can be raised. 6.14 Fiscal performance can be a good indicator of the economic potential of an urban entity. It could be assessed through indicators such as the fiscal uotential (measured, for instance, by the number of inhabitants and the number of taxpayers actually paying taxes), and the fiscal effort (measured, for instance, by the ratio of total fiscal revenues to the number of taxpayers). Half of the transfers could be allocated in this way, which would give more money to entities with a gocod economic potential and reward local governments which make an effective fiscal effort. The other half could be allocated according to infrastructure needs as measured by such indicators as number of pupils, number of houses and population density. 6.15 This proposal for new al"ocation criteria could serve as a basis for the reassessment which will be required. A joint initiative should be launched to redefine the allocation criteria. It would associate Central Government (Ministry of Finance, Ministry of Interior, an' the technical ministries, particularly Education and Health) and representatives of local authorities, who could be chosen among the members of the Association of Presidents of urban entities. B. Local Taxes 6.16 Fiscal orzanization. The Central Government is responsible for assessing and collecting most local taxes. Two departments play a crsucial role in this respect: the Department of Taxation, and the Treasury Department. The Department of Taxation, in the Ministry of Finance, determines the tax base for local taxes in addition to all other dire:t taxes. Collection is carried out by the Treasury Department. 6.17 Revenues from local taxes have been steadily decreasing as a percentage of Central Government revenues. This is the result of a combination of factors such as the concentration of the responsibility for collection of local taxes in the hands of the Central Government, the lack of coordination between the different Departments concerned and the lack of adjustment in the level of local taxes, in what has traditionally been an inflationary environment. Central Government revenues, which mainly rely on indirect taxation, have been better shielded. There was also a general deterioration in the performance of Central Government with respect to tax base determination, tax computation and tax collection. Moreover, taxpayers' 8ttitudes have been changing. The deterioration in the quality of public - 64 - services and in the general economic situation has eroded taxpayers' willingness and capacity to pay. 6.18 The general lack of coordination between the various agencies involved in assessment (Department of Taxation), collection (Treasury Department) and management (local governments and Ministry of Interior) of local government resources is at the root of the problem. The Ministry of Finance considers local taxation as a burden. Local taxes are difficult to assess and collect, and the proceeds are turned over to others. The Ministry of Interior is much more concerned but is not always well informed of the exact situation of local finances. No one today has a general overview of local finances. It should be recognized that any improvement in local resources will require the cooperation of a number of entities (Ministry of Finance, Ministry of Intexior, technical ministries, etc). Therefore, there is a need for an aggregate approach, going well beyond the confines of local government. 6.19 Given the national macroeconomic objective of reduction of fiscal pressure, the main actions to be taken are to enlarge the tax base to reach people who presently escape taxation, and to im rove the effectiveness and efficiency of tax colletion. Repartment of Taxation 1. Tax base determination 6.20 Property and land taxes are difficult to determine because the Department of Taxation knows little of the real situation, as far as property is concerned. One employee is often responsible for as much as 5,000 Items. The actual number of houses is very often larger than the number of houses included in the Department's records. Many familiea and businesses escape any kind of taxation. Annexes and other dependencies of properties of less than 2,000 m2 In area are exempted from the land tax. Thus, a dependency of 1,900 m2 and of 200 m2 are treated fiscally in the same way. This is a cause of tne low yield of the land tax (about FMG 13 million per year in Antananarivo). The yield of property and land taxes could be significantly improved through the implementation of the following measuress (i) tstablishment of a simplified land register, based on aerial photographs. This would be a great help for the agents responstble for determining the tax base; (ii) better use of the information already available: water and electricity files, knowledge accumulated by the agents, and registers of the different levels of local authorities, especially Firaisana and Fokontany. In particular, the Fokontany know their constituencies well and by and large have little problem collecting the "social contributions' which constitute their main resourcel and - 65 - (iii) use of a more differentiated approach in urban areas, taking more account of building area, quality of available infrastructure, etc., in order to make a better assessment of the taxable items and of the actual capacity to pay. In particular, for the land tax, land area should be better taken into account and the present taxation threshold (2,000 m2) should be lowered to 150-200 m2. 6.21 For the business tax, in a similar fashion, the approach should involve a broadening of the tax base rather than an immediate increase in rates. One part of the tax is proportional to the rental value of business premises. For that part of the tax, the above recommendations on property and land taxes fully apply. 6.22 The informal sector benefits from urban infrastructure, and should therefore also pay the business tax. Nevertheless, this should be done with care, since one source of the informal sector's economic power is the very fact that it is informal and therefore escapes government intervention. Any tax introduced should be simple and fixed at a low level. It bhould also be a standard tax to eliminate uncertainties on its computation. Finally, its payment should be easy to control. For instance, a receipt could be delivered to all taxpayers. Such a system would greatly reduce the possibilities of tax evasion and of negotiations between taxpayers and fiscal agents. 6.23 The Faritany are today the only local authorities which benefit from the business tax, whereas the Fivondronana and Firaisana are actually responsible for providing businesses with a suitable infrastructure. Therefore, a significant part of the business tax proceeds should be transferred to the urban Fivondronana and Firaisana. This action would significantly improve their financial balance and enhance their capacity to finance essential infrastucture. 6.24 If attention is devoted to the above issues, it should be possible to increase local tax revenues significantly over the next several years, without increasing individual tax burdens. The example of Toamasina, where local tax revenues increased by 67 percent from 1987 to 1988, demonstrates that this can be done, at a reasonable cost and in a relatively short time. 2. Tax computation 6.25 Property taxes are based on voluntary statements. For rental properties, landlords have to report their rental income to the Department of Taxation. For owner-occupied properties, this Department assigns inputed rents. In fact, only 25 percent of registered landlords, and an even lower percentage of all owners, actually make a statement. Moreover, these declarations generally underestimate the rents actually received and the rental value of the property. For example, according to owners' statements the rental income received from two similar buildings in comparable - 66 - districts of Antananarivo is supposed to vary in a ratio of 1 to 9. In addition, there is little or no ex-post monitoring of landlords' declarations. Another cause for the low yield of property taxes has been the interval (5 years, recently reduced to 3) between adjustments, which mskes it difficult to keep pace with inflation. The lack of a smooth adjustment process for assessed property values has led to abrupt one-time adjustments, which meet with strong resistance. The first Urban Project (PDUM, Credit No. 1497-NAG) has made a start at improving this situation, for example by financing periodic assessments and strengthening the capacity of the Department of Taxation. However, this process is still in its early stages, and this improvement program will continue to need strong support for several years to come. 6.26 The Central Government should establish simple and uniform criteria of computation of the rental value of property. These criteria could include factors such as the quality of building materials, type of building, and access to basic infrastructure. This would lead to the computation of an index defining a rental value per square meter for each specific urban area. The index could be revised each year to reflect trends in the underlying factors. Finally, the landlords' declarations should be very closely monitored, and delays and inaccurate statements should be effectively and rapidly sanctioned. 3. Tax reaisters 6.27 The register of taxpayers for the property tax Is often issued late. In Antananarivo, by the end of 1987, that delay had reached two years. Several Goverrment units, including the Divisions of Direct Taxation, Indirect Taxationg and Registration, intervene in the process, which creates a number of coordination problems. Moreover, the registers are often inaccurate and errors can be carried forward over several years. 6.28 The recent progress made in Toamasina demonstrates that improvement is possible, at a reasonable cost. The first priority is to catch up with the delay already accumulated and meet the schedule for issuing future registers. For this to be done, a number of conditions need to be met: - the present computer system which does not allow updating of data and is not interactive must be redesigned. It must become possible for the departments which use the information to have direct access to the system database and to update it. The computerization of register management should also be extended to the main city of each Faritany; - training must be given to the teams which will be responsible for the system; and - 67 - - the Fivondronana or Firaisana should be asked to finance a part of the budget of these divisions. Treasury Department 1. Tax collection 6.29 According to the rules of public accounting, the Public Accountant ("Comptable public") In each city or town is solely responsible for tax collection. But the decree defining the reaponsibilities of the Presidents of Executive Committee states that the President of the Executive Committee is responsible for tax collection and has authority over tax collectors 24/. 6.30 The absence of a clear definition of responsibilities for tax collection creates needless confusion. To avoid political difficulties, most Public Accountants leave the responsibiiity for tax collection to the Delegate of the Administrative Committee, limiting their action to acknowledging the transfer of the taxes collected by local authorities. But the Administrative Delegate Is generally more sensitive to political pressures than the Public Accountant and Is often less able technically. Moreover, many towns do not have an Administrative Delegate. This situation leads to very low collection rates (between 10 and 15 percent of the tax bills sent out). 6.31 The first step to improve this performance should be to better define the responsibilities of Central Government and local authorities. The roles of Public Accountants and elected officials should be more precisely defined. It is recommended that Public Accountants shoulds - plan and organize the tax collection visits and the work of the tax collectors; - be personally and financially responsible for all taxes due until they are either collected or declared unrecoverable; and - for all unrecoverable taxes, be able to demonstrate that due diligence has been used. Elected officials shouldt - prepare the tax collection visits locally by making the schedule of these visits known and by working at the local level to make taxpayers better aware of their obligations; 241 Decree 77-413 of December 22, 1977, Article 2. - 68 - - take part in the distribution of warnings to bad taxpayers; and - ensure the physical security of funds and of tax collectors. Presently, the responsibilities for issuing the tax bills and for monitoring the results of tax collection are separated. Local governments' accounting units should be made responsible for both tasks, so that they can have a broad view of the effectiveness of the collection process. Local governments must play a much more active role in resource mobilization than they do today. It should be noted that the slowness of procedures and the length of tax recovery period are not the only causes of poor receipts. Other factors such as inadequate resources for collection, tax evasion and fraud, and indifference among local officials also need to be addressed. 6.32 A number of positive steps are being taken by the Central Government. For instance, the 1989 Appropriation Law includes a reduction from one year to three months of the time after which a fine can be imposed on a faulty taxpayer. That should make the Public Accountant's task easier. But some other issues require clarificati,n. For instance, there are no clear rules governing the granting of grace periods to taxpayers; the situation therefore varies widely from one town to another. Central Government, particularly the Department of Public Accounting, should prepare a hand-book for use by Public Comptrollers ("Controleurs Financiers") in order to clarify such issues as the rules governing the granting of grace periods to taxpayers, the handling of taxpayers' complaints, and the follow- up to these complaints. 2. Bad debts 6.33 For the actions suggested to be fully effective, the accounts of local authorities must be rectified by writing off the bad debts wkich have been carried forward from previous years. These bad debts represent more than FMG 10 billion and are often treated by local officials as a potential source of revenue. In fact, most of them are too old to be ever recovered. Moreover, they increase the administrative burden of the Treasury Department. 6.34 The first steps toward a solution should be to establish in each municipality a list of all debtors and then define three categories of debts, namelyt (a) more than three years old; (b) between one and three year. old; and (c) less than a year old. Category (a) files should be kept by the Central Government and a special procedure should be adopted by which they would be written off as a one-time exercise. Elected officials and Public Accountants, informed of the write-off, would then focus their efforts on recovering the debts in categories (b) and (c). - 69 - 3. Treasury Department budgets 6.35 There is a strong correlation between the resources available to the Treasury Department and its effectiveness. In a number of cases, the budget does not even cover fixed costs such as water and electricity. Local authorities, who benefit from the action of these departments, should contribute more to the financing of Treasury budgets, as was suggested for the Department of Taxation (para. 6.28). For instance, the Fivondronana of Antananarivo-Renivohitra has supplied 10 motorbikes to the tax collection division. That example should be followed by other Fivondronana and Firaisana. Finally, the ceilings currently imposed on the financial incentives received by tax collectors should be lifted, in order to improve their motivation and to make them less tempted by corruption. C. Other Revenues 6.36 A portion of local government revenues is directly controlled by elected officials. These revenues fall into two categories: revenues from goods and services, and parafiscal revenues. 1. Revenues from aoods and serv' as 6.37 Even though local governments do not rely heavily on them today, revenues from goods and services have a strong potential. They consist of three main categories: (i) revenues from private property (rents, etc.); (ii) revenues from public property (market stall fees, water and electricity taxes, etc.); and (iii) revenues from services (septic tank draining, fire brigades, etc.). When the reluctance of Popular Councils to raise any taxes or fees can be overcome, each of these categories can become a major resource for local governments. Local governments should focus their efforts on revenue sources whose base can be readily identified and which are easy to collect (especially rents and market stall fees). 6.38 The Fivondronana of Toamasina provides an interesting example in this respect. In 1988, the Fivondronana increased its rents by more than 20 percent. In addition, a tax of 7 percent was imposed on the cost of each Kwh and m3 consumed. Together with improved tax collection, these actions resulted in a 66 percent increase in revenues from goods and services in Toamasina in one year, from FlMG 129 million in 1987 to FMG 214 million in 1988. Opportunities for increasing revenues need to be explored on a case- by-case basis, >Ince there are significant differences among municipalities. As one illustration, Toamasina owns more than 4,000 properties, Antananarivo only a bit more than 300, and Antsirabe none. In Mahajanga, which also owns a large number of properties, rents have not been adjusted in more than 20 - 70 - year.. Revenues from goods and services can be significantly Increased in most municapalities. 6.39 Significant gains can also be achieved in market stall fees. Tax base determination and tax collection need to be Improved. In most cities, it would be worth strengthening or eatablishing a Markets Department. Its main functions would be to monitor the ssignment of stalls, register merchants and collect taxes. It would also asi at controlling the physical expansion of cosmercial activity and preserving public land. Market stall foes often represent as much as lU percent of current revenues. Conparisons with other countries suggest that they could constitute a major resource for local government (21 percent of local tax revenue In Dakar), thereby justifying the Initial costs incurred to improve the assessment and collection system. 2. Parafiscal revenues 6.40 Revenues of this type consist of a variety of taxes and fees on agricultural/forest/mineral products, alcoholic beverage sales, shows, pin- ball machines, advertising, slaughtering, and the like. A number of these taxes are linked to the fact that several urban entities still have rural areas included in their territory. The level of most parafiscal resources wse determined more than 10 years ago and has seldom been adjusted since taen. Today, they understandably represent a nominal portion of local government revenues (generally between 1 and 5 percent). These taxes are difficult to collect and share a number of characteristics: the staff responsible for determining their base and collecting them has little or no training; the importance of personal relationships makes recovery by local officials arduous; and it is very difficult to take any meaningful legal action against most of the potential taxpayers, who have few assets and feel little sense of obligation towards urban entities. Finally, the law creating those taxes 2S/ was never follorwed by an application text. Therefore, the type of sanctions which can be enforced is unknown. 6.41 Given these characteristics, local authorities should aim only at streamlining and rationalising the management of those taxes. The first principle should be to abolish all taxes whose collection cost is greater than the revenues generated. Local governments should focus their efforts on the *best' taxes, i.e., with the most favorable benefit/cost relationship. Finally, their level should be adjusted on a yearly basis, to avoid large one-time increases. £l Ordinance 78-010, of May S, 1978. - 71 - FINANCIAL MARAGEMENT 6.42 Efforts devoted to securing more resources for local governments will have been wasted unless financial management at the local level is significantly improved. Local governments will have to enhance their planning and budgeting skills in order to establish realistic forecasts. This will require a better understanding of the legal framework and improvements in budgetary procedures. Local governments will also have to improve expenditure Programming and control so that expenditures are kept within bounds. This in turn will require a number of changes in management information systems, planning and budgeting tools 8nd staff skills. A. Le2al Framework and Budgetary Procedures 6.43 The legal framework consists of four main texts on local financial managements (i) Ordinance 78-010 of May 5, 1978, which defines the nature and the base of the taxes available for local governments; (ii) the General Tax Code linked to Law 77-005 of December 12, 1977; (iii) Decree 68- 080 of February 13, 1968 defining the rules of public accounting in Madagascar; and (iv) the yearly appropriation laws. These texts sometimes lack precision and allow for diverging interpretations. There is no comprehensive text describing on a practical level how the legal requf-3ments should be applied. Moreover, these texts are generally not available locally and when they are, most employees are not able to interpret them. In addition, the legal framework is not consistent with the institutional reorganization of 1976 (Ordinance 76-044, of December 27, 1976). For instance, the budgetary nomenclAture for rural entities differs from that for "ex-communes urbaines", which hinders analyses and statistics on a national basis. 6.44 The situation can be readily improved if the Central Government, particularly the Ministry of Finance, takes action. First, a 'Code of Local Governments", grouping all the relevant legal texts and their accompanying measures, should be issued. Second, a 'Financial Management Guide' should be drafted, with a detailed description of the main procedures, the purpose of the different forms and procedures, the main deadlines, and the links with Central Government. Finally, the accounting nomenclature should be streamlined in order to make it consistent with the chart of accounts and to prepare the future introduction of stock and cost accoun_.ing. The accounting nomenclature should allow for differences between rural and urban entities without sacrificing the ability to produce national statistics. 6.45 The requirements for budgetary procedures are different for each level of local government (see Annex 6.2), but the comments made here are valid for all local government levels. The analysis of the draft budgets ("budgets primitifa") shows that there is generally little correlation between forecasts and previous results. On the one hand, revenues are often overestimated; it is not rare to see revenues barely exceeding FMG 10 - 72 - million being budgeted year after year as WMG 100 million, with the approval of the supervising authorities. On the other hand, expenditures are often underestimated, leading to the accumulation of arrears at the end of t;.e financial year. The systematic overestimation of revenues and underestimation of expenditures almost inevitably leads to deficits. Moreover, there is no revision procedure which would allow for a modification of the forecasts during the financial year. The 'Additional Budget' is nothing more than the recognition and integration in the new budget of the previous year's surplus or deficit. The problem is compounded by the absence of a fail-safe mechanism which would trigger action when the deficit exceeds a given level. Last but not least, budgets are often put in place late in the financial year, due to the slowness of administrative procedures. 6.46 The quality and the utility of budgetary procedures could be improved by implementing several simple measures. Every year, in September for instance, the Central Government should issue a methodological note defining budgetary directions and targets (financial ratios, thresholds), in order to help local authorities achieve a better financial balance. The supervising authorities should also define a compulsory timetable and reduce the time limits for budget approval. The rule should be that a budget would be considered approved, if no response had been received by the end of the review period. 6.47 The budgetary process should also be made more flexible. Local governments could modify their budget in reaction to major changes in the economic environment. A procedure should furthermore be set up to automatically call for Central Government intervention in case the deficit exceeds the financial capacity of a local authority. Central Government should define and make known to local governments specific thresholds based on simple ratios such ass actual expenditureslforecast expenditures, actual revenuesiforecast revenues, etc. But these improvements in budgetary procedures and control will bear fruit only if a parallel improvement in expenditure programming and control takes place. B. Expenditure Proaramming and Control 6.48 Since the beginning of the 1980s, total expenditures have barely increased in current terms, whereas cumulated inflation since 1980 is over 200 percent. This evolution has led to a dramatic decrease in real expenditures, particularly for capital improvements, which is now jeopardizing the future of Malagasy cities. Expenditure programming, particularly the proper balance between current and capital expenditures, and expenditure control must be Improved to avoid the 1990s resembling the 1°80s for Malagasy cities. - 73 - 1. Current Expenditures 6.49 In 1984, the Fivondronana were spending on average FMG 228 per inhabitant per year. Given the scarcity of their resources, it is of utmost importance for local governments to have a tight control over their current expenditures. It seems neither possible nor desirable today to redtuce total current expenditures. One cannot expect any substantial increase either. Therefore, the main issue here will be to modify the structure of expenditures so that they are devoted to priority needs. This requires an analysis of the main types of expenditures: personnel, weter and electricity consumption, waste collection, sanitation and road maintenance. These expenditures represent over 80 percent of the total. 6.50 Personnel expenditures represent on average more than 60 percent of local governments' current expenditures. It is the biggest single expenditure. The erosion in local government resotrces over the last decade was not followed by a parallel reduction in personnel expenditures. On the contrary, as a result of social and political pressures, they often Increased, generally at the expense of other current expenditures and of capital outlays. In Antananarivo-Renivohitra, for instLnce, personnel expenditures increased by 40 percent from 1983 to 1986 while capital expenditures remained constant in current terms. Toamasina, where the workforce was reduced from 665 to 400 people over a three-year period, is one of the few exceptions. Salary growth in the future should be strictly controlled and no real salary increase granted without a corresponding increase in productivity. 6.51 The Toamasina example notwithstanding, firing employees in order to hire new ones with more suitable skills is not feasible on a national scale. Local governments must do better and more with the same staff. At present, there is a mismatch between the skills of the workforce and those actually needed by local governments. In particular, intermediate-level staff are in short supply while unskilled workers are abundant. This is illustrated by the high ratio of drivers to vehicles and secretaries to managers. There is consequently a need for a comprehensive ManRower Development 2lan for local government staff. 6.52 In effect, the availablity of a large workforce is an asset for local governments if it is well managed. This will have to be done through a major training effort and through reallocating staff to new tasks. Personnel skould be allocated in priority to revenue-generating tasks. Job desiriptius should be prepared and a training program established to better match staff skills with local government priorities. The job descriptions would also define performance criteria that would help make promotions more trasparent and less dependent on political considerations. Internal control should be tight and sanctions in case of wrongdoing should be strictly applied. Significant savings can be achieved by insisting on punctuality and effective presence at work during working hours. In effect, because of the present level of absenteeism, a number of local governments must have more workers than they actually need. The new human resource manag'ment policy - 74 - should be to have fewer people, but better paid, better trained, more motivated and more productive. 6.53 The Fivondronana are supposed to pay for water consumed at standpipes. They are generally unable to finance that cost with their present resources, especially since the recent increase in JIRAMA tariffs. For exaAple, In 1987, the water consumption bill for the Fivondronana of Antananarivo-Renivohitra was FMG 1 billior., which represents 30 percent of its current expenditures. Arrears prior tc December 12, 1986 were paid by Central Government, but the Fivondronana of Antananarivo-Renivohitra had already owed FMG 2.5 billion to JiramA for 1987 and 1988. Because of a high population growth rate and of the density of urban settlement, water consumption expenditures of local governments are rapidly increasing. The problem is compounded by the fact that JIRAMA bills are often inaccurate, billing local governments for water losses for which they are not responsible. Much progress has been made in the long-standing Issue between the Government and the Fivondronana of Antananarivo-Renivohitra (see para. 2.13). However, given the urgency of new water investments in A tananarivo, the final resolution of this issue needs to be achieved without delay. 6.54 The preferred solution In the long run is to encourage Individual connections to the distribution network. In the meantime, a number of actions should be taken by local governments in order to avoid accumulating arrears: (A) installatior of water meters at all consumption points, particularly standpipes and wash-houses; and (ii) use of municipal employees to track the consumption and prepare statistics on the consumption points (average weekly consumption, etc). This would allow comparisons between different points and would make it easier for local governments to identify overconsumption (water vendors, small business consumption) or overbilling by JIRAMA due to leaks. Consumption monitoring is an area where reallocation of personnel could prove quite beneficial (see para. 6.52). The Fivondronana of Antananarivo-Renivohitra has started to monitor its consumption and seems to have achieved savings on the order of 5.) percent i.e., FMG 600 million per year. 6.55 In any case, it will take a long time before the number of individual connections is high enough to have a significant impact on local authorities' water bills. This further strengthens rhe case for an effective cost recovery system. There are three main options: (i) rent the standpipes or wash-houses to individuals or firms which would manage them according to contracts signed with JIRAMA. The individuals or firms would pay the water bill and be reimbursed through charges for water consumption. Such a system has been successfully implemented for the wash-house of the Firaisana of Ambodimanga in Toamasina; (ai) establish paying standpipes, as is done in Cote d'Tr'oire ar.d Gabon, incurring the higher initial cost of this ecjLpment; and (iII) achieve indirect cost recovery through increases in municipal revenues to cover water expenditures. - 75 - Each local government should adopt the solution best suited to its needs and situation. The cost recovery system adopted should aim for balance between receipts and expenditures, not only by reducing expenditures through lmprovements in the management of resources, but also by encouraging discipline in the payment of tax bills. 6.56 Electricity and water consumption raise the same type of issues. The existence of illegal reonnections, for instance In the markets, makes the situation even more complicated in the case of eleetricity. Most local authorities are in arrears and cannot pay their electrieity bills. The situat£(on is likely to worsen since public 14ahting needs to continue to expand with urban growth. The most urgent - lon 's to carry out a complete inventory of existing electricity meters. Following this Inventory, all 1i1.egal connections should be suppressed. An effective monitoring system should be established in order to track consumption and identify potential revenue liss from illegal connections. Local authorities could organize a card-file to identify all meters for which they are responsible. Employees could be reallocaced to that task (see paragraph 6.52). The savings achieved would, without doubt, be very significant. 6.57 Unlike water, public lighting is not conducive to direct cost recovery. Indirect cost recovery (through municipal taxes) must be applied. Therefore, all necessary measures should be taken to tailor public lighting investments to the real financial capacity of each local government. It is appropriate, for example, to ask whether public lighting should be introduced into areas in which most residents do not pay their local taxes. Cities and towns should also look for creative ways to reduce their electricity bills, such as not turning on every street lamp or reducing hours of coveraSge. 6.58 Local governments are responsible for solid waste collcction. This operation is generally costly, given the number of workers mobilized and the nature of the equipment used. In the Fivondronana of Antananarivo- Renivohitra, the overall cost of these activities is estimated at about FMG 715 million in 1987 (PMG 465 million for wages and FMO 250 million for fuel, machine maintenance, spare parts, and so on). This amounts to FMG 1,036 per inhabitant per year, i.e., 33 percent of current expenditures. The "Additional Tax on Built Property" (Taxe Annexe a l'Impot Foncter sur la Propriete Batie, TAFB) is supposed to finance those expenditures but its revenues are far from sufficient. 6.59 The service provided is generally of poor quality. This has an impact on collection vf some taxes, such as the market tax. Since there is no cost accounting system, the cost of waste collection is unknown, encouraging an overly passive attitude among elected officials. The first priority should be to set up a simplified accounting and budgetary control system which would provide preliirinary estimates of several ratioss - volume of waste collected/number of workers; - maintenance costsldepartment budget; - 76 - - personnel expenditures/department budget; - tax revenues/department budget. These data could be centralized by the Ministry of Interior. They would then be transmitted back to local governments, which would be able to use them as performance criteria until a true cost accounting system can be put in plac4. Local governments should also explore all avenues to reduce their expenditures and increase their revenues. For example, animal traction could be used as a substitute for mechanized equipment in the suburbs. Collected waste could also be used commercially. 6.60 Sanitation and road maintenance are probably the most difficult tasks for Malagasy local authorities, due to the scarcity of resources. These services are capital-intensive. No major investment has been made for years and the existing infrastructure has not been properiy maintained. The recent return of diseases like cholera and malaria is a consequence of that situation. Lack of a proper sanitation network is very costly and Investment would generally have a high return. Therefore, improving the sanitary environment shoud now become a top priority for local authorities. A number of steps should be taken: - local authorities should not only accept but also manage the existence and the development of spontaneous settlements; and - they should also prepare multiyear development plans with the help of technical ministries and private firms. These plans would anticipate and accommodate urban growth by providing for basic services in new settlement areas. 6.61 Law 88-013 of October 25, 1988 has clearly defined the distribution of responsibilities for road maintenance between the central and local governments. The "ex-communes urbaines" are now responsible for maintaining their roads. Maintenance costs will be partly financed by the Road Fund which has now been operational for two years, but part of the funds will also have to come from the towns' own resources. Local governments should use private contractors and local technical departments (Services Regionaux des Ponts et Chaussees, SRPCH) to carry out maintenance work. Lit, in any case, little will change unless significant and sustainable improvements in resource mobilization take place. 6.62 Local governments have often accumulated arrears of several years in contract payments, with the result that they are not trusted by suppliers. It is therefore not rare for them to be charged prices up to 40 percent above market levels, for example for spare parts. In order to re- establish local authorities as reliable customers, accumulated arrears will need to be paid off. This will require joint financing by locel authorities and the Central Government. This has already been done successfully in Antananarivo, and could be repeated elsewhere. Through this means, a temporary improvement in the financial reputations of towns can be achieved. Whether this improvement is sustained in the future depends once again on the resources flowing to local governments. - 77 - 6.63 All activities and services currently performed by local governments should be scrutinized and compared with the service provided by the private sector, in terms of quality and cost. Such an analysis could lead to additional savings and has already been completed for a number of sectors such as waste collection, sewerage, and urban drainage for Antananarivo. 2. Capital Expenditures 6.64 Capital expenditures, which in the 1970s represented a substantial portion of total expenditures, have sharply decreased. This is confirmed by the visible deterioration in the Malagasy urban infrastructure, which, if it continues, would jecnardize the production potential of the national economy. This is the consequence of a lack of resources but also of a lack of project identification and implementation skills. It has now become critical to increase capital expenditures in the cities and towns. 6.65 All local government revenues taken together are not sufficient to cover current expenditures and develop a satisfactory self-financing capacity. The local tax baqe will not be sufficient to finance the scale of investment needed to accommodate the urban growth of the 1990s. The unavailability of external financing (bank credit, etc.) for local governments, due principally to their lack of creditworthiness, accounts for much of the present underfunding of urban areas. Specisl mechanisms suited to the nature of local authorities and to their projects should be established. The Fund for Infrastructure and Urban Housing (Fonds .d'infrastructures pour l'habitat urbain, FIHU) could play a central role in this respect. FIHU was created to receive the repayments linked to the subsidiary loan of the first Urban Project as well as funds from other sources. Its objective is to channel those funds to the urban sector. During the first five-years, FIHU operations would be managed by a private bank. However a coordinating committee comprising all agencies Involved in the urban sector would set policies and rules for the activity of the FIHU. At the end of this five year period, FIHU could be transformed into an independent infrastructure bank. The FIHU should have an incentive function built into its operations. In effect, the Central Government could include a degree of conditionality in the allocation of funds, in order to reward urban entities which will have engaged in a genuine financial management and renouree mobilization effort. 6.66 Despite the scarcity of their resources, a number of local governments have unspent funds at the end of each fiscal year. This is due to a lack of project identification, planning and implementation capacity. Ldocal authorities should also be provided by the Central Government with technical support for project analysis and project implementation, or should be able to acquire tbat service from the private sector. In any case, local government staff should be trained in order to acquire the basic planning and implementation skills without which any increase in capital expenditures would be useless. - 78 - 6.67 These changes would enable local governments to design carefully prepared investment programs, consistent with local needs, with their increased self-financing capacity and implementation skills. Local governments could then sign proRram-contracts with the Central Government, committing themselves to achieving a number of financial and operational targets. External financing would be available only to local authorities satisfying a number of conditions, i.e.,: - generation of stable revenues anl surpluses in a sustainable manner; and - investment program-ing consisting of projects selected and evaluated according to economic and financial criteria. This approach would favor projects with high rates of return, which would enable the local authority to reimburse its loans safely. These projects should be consistent with the objectives of the liberalization programs now underway, i.e., - strengthening financial balance ard reducing the drain on the Central Government budget; - creating a suitable enabling environment for business; and - encouraging and iromoting export-oriented production. 6.68 With regard to capital expenditures, one of the key concerns Is undoubtedly the sharing of the burden between central and local government. This question. needs to.be.oaddressed in a long-term perspective. Ultimately the main responsibility for local expenditure wil1 be with local governments themselves. Even then, the Central Governmeac will play a critical role in urging CDs to state their priorities clearly, program expenditures accordingly, and monitor them effectively. In the near term, the Central Government must not only foster these roles among CDs, but carry them out itself for the local expenditures it undertakes directly. 6.69 Any solution should be based on a phased approach. From the financial standpoint, a sliding scale should be developed whereby the Central Government would indeed provide significant resources to cities, but at a decreasing rate over the program-contract period. An action plan for capital expenditures could be divided into three five-year phases: Phase Is Major Central Government involvement, financially and technically, would characterize this phase. Local governments would sign proxram-contracts with the Central Government, defining yearly objectives and the means available to achieve the,.t. Phase 1 would aim to Improve local governments' financial situation and build their capacities. - 79 - Phase 2: Local governments would be more financially self-sufficient and better managed. Phase 2 would be devoted to a major reconstruction effort for which local governments would use FIHU, banks, and other external financing. Phase 3t Local governments would be able to Zully play their role in the Malagasy economy and be genuine engines of growth. Urban growth would take place in a much sounder environment and would be better linked to national growth strategies. 6.70 For this ambitious picture to become reality, a number of conditiona will have to be mett (i) Central Government will need to have the political will to decentralize and to fully accept the consequences of decentralization. Collaboration and coordination between Central Government and local governments will have to improve and become more transparent; (ii) Local governments will need to have the golitical will to engage in a resource mobilization effort which will not be popular. They will also have to be willing to work together with Central Government and its local representatives; (iii) A major capacity building effort will have to take place at the local level. Carefully designed and implemented ttaining programs will make the staff more effective and efficient. Better human resources management (para. 6.52) will enable local governments to attract, motivate and retain highly qualified staff; and (iv) Planning, implementation, and monitoring will require improvement, as described earlier in this chapter. - 80 - CHAPTER VII URBAN LAND MANAGEMENT A. Land for Urban Growth: Concept and Reality Setting 7.1. Even though the rate of urban population growth is moderate, urban areas have been unable to provide buildable sites in the numbers required. It Is estimated that in Antananarivo, the housing stock increases by 3,000 units a year but that at least 6,500 units would be required each year to accommodate the demand -see also Annex 8.1). Of the 3,000 units built annually, 851 are being constructed informally, in large part because of the scarcity of buildable land. 7.2 Urbanixation, the poverty of large numbers of households, and the shortage of municipal resources are the most critical issues affecting land mansgement practices. In this context, ocacupation of land without full title and at low housing standards represents the only alternative. The problem, as usual, is more complex and therefore more serious in Antananarivo than in other cities of Madagascar. Topography, the scarcity of land protected from flooding, and the absence of public land reserves combine to create an extreme scarcity of land for urban uses. According to the Greater Antananarivo study, only 1,000 ha of the 4,700 ha of State-owned land are located in the Greater Antananarivo (GA) region 26/. Of this amount, only 400 ha, fragmented into about 40 parcels, lie within 10 kms of the city center. According to the same study, more than 75 percent of all parcels in GA are less than 1,000 m2 in size, demonstrating the extreme level of land fragmentation in urban areas 27/. 7.3 Private land markets are not very active. The number of formal land transactions among individuals is limited, although transactions between families can be numerous. Even in the towns, the Malagasy practice of holding onto land for the next generation is still strong, especially in the Merina tradition. 26/ AURA/Groupe Huit/BCEOM, Etude de Developpement du Grand Antananarivo, 1985. 27/ The figure remains at 671 even when the heavily rural Fivondronana of Arivonimamo is Included. See Annex 6.7. - 81 - 7.4 The following sections will focus on three main groups of issues affecting land management and the functioning of land markets in cities and towns. The issues are (a) land rights and the existing registration system; (b) development control and guidance; and (c) issues arising from the settlement of urban land. The concluding section describes the most important implementation issues and outlines a three-stage progran for reform of urban land management. Land Riahts and the Registration System 7.5 Existing legislation is an impediment to regularization of land rights. Urban land, accordiag to this legislation, can be granted full title only if it was occupied prior to February 15, 1950 and if it has been improved (mis en valeur). Since only a handful of people can show evidence of continuous occupation over a span of 40 years, this stipulation greatly limits the regularization of urban land and contributes to the growth of unplanned settlements. An alternative means of acquiring clear property rights is through successive possession (adjonction de possession), in which the current occupant, even though not on the property since 1950, can show who the successive occupants were since then. 7.6 According to Law 90-028 and application decree 90-656, both of December 12, 1990, registration of property is required. The registry books and cadastre are updated on the basis of registration of transactions by the Land Registry Bureau (Service des domaines, SD). However, this system cannot deliver accurate information on land ownership, for two main reasons. First, because of the high costs of registry, many transactions and changes are not recorded. This means that the boundaries between public lands, private state lands and private lands cannot be determined. 28/ The Government does not know how much land it owns. Second, there is little collaboration between units responsible for land registry and for land taxation. The Property Directorate (Direction du patrimoine, DP, in the Ministry of Agricultural Production and Land Resources, MPAPF; SD is a part of DP), does not communicate well with the Property Tax Bureau (Service des contributions directes, SCD, in the Finance Ministry). For example, the computerized data collected by SCD through periodic censuses of property are not transmitted to DP. 28/ In Malagasy law, public land is of two types. "Public lands" (domaine public de l'Etat) are the roads, parks, and the like which are administered by DP for the nation as a whole, are non- transferrable, and on which no private development is to take place. "Private State lands" (domaine prive de l'Etat) are public properties registered in the name of a particular Government ministry or public agency. Private State lands can be transferred from one Government unit to another for specific purposes in the public interest, such as a housing project. Generally, Government is better equipped to deal with occupation of private State lands than with squatting on the dikes of the Antananarivo rivers and other public lands. - 82 - 7.7 This situation leads to many different shades of tenure arrangements. Private, registered property occupied by the owner is the only fully legal tenure situation, but it applies to less than 10X of properties. For urban land occupied and improved before 1950, and rural land occupied for 10 years, legal ownership is not automatically granted but can be if the beneficiary wishes it. This too applies to only a small group. In the vast majority of cases, people have biilt h. ;ses in cities and towns since 1950 on land that is either of uncertain owaership or registered in the name of someone else. Housing built on private registered land but belonging to another owner, or on private state land (domaine privi de l'Etat) is clearly illegal. So is housing on land whose tenure status is undefined, since that is presumed (unless contested) to be private state land. 7.8 One would expect, in this context, :hat as much as 85S of new construction is carried out without building permits and formal registration procedures. The magnitude of the problem is such that new ways should be introduced to regularize a situation that affects most urban residents. Development Control and Guidance 7.9 Madagascar is experiencing problems typical of many developing countries in establishing an appropriate institutional and regulatory framework for land development. Existing master plans (sch&eas directeurs) date back to the 1960s (Antsirananat 1963) or early 1970s (Antsirabe: 1971). Although these master plans have never been approved, they have constituted the basis for existing land use regulations. Three master plans are presently being prepared for the cities of Antsirabe and Mahajanga, and an indicative plan was recently completed for Antananarivo. These documents however would have no statutory power, but are being complemented by land use/zoning plans which have been under preparatlon for several years but have not yet been approved. The Urban Planning Code (Code de l'urbanisme) recommends the application of a set of sophisticated urban plans, including "plans d'urbanisme directeurs," "plans d'urbanisme de detail," and "plans somamires." Plans can be approved only after a complex and time-consuming review process by the Council of Ministers 29/. 7.10 According to the Code de l'urbanisme, whoever wants to build in a town of more than 2,000 inhabitants must obtain a building permit. The request is reviewed by the regional branch of the Directorate of Architecture, Urban Planning and Housing (DAUH), of the Public Works Ministry, and is granted by the President of the Executive Committee (Mayor). After completion of the works, a "certificat de conformit;" is granted. Such a system permits both the national and local authorities to oversee the construction process. However, the application of these procedures is very limited. Only about 300 building permits in Antananarivo, and 20 in secondary cities, are processed each year. The 291 Articles 19-28, Code de l'urbanisme. - 83 - absence of urban plans as reference documents, high and costly construction standards, and weak municipal technical services are the main reasons for this limited application. 7.11 Subdivision procedures as outlined in the Code de l'urbanisme are complex and highly centralied. Authorization is given only by the Ministry of Public Works following studies undertaken by the Ministry. The developers themselves must provide Infrastructure. Settlement of Urban Land 7.12 These features of the legal framework for urban land management leave a wide scope for families to occupy virtually any land without prior authorization. Families may submit a formal request for a piece of land and believe that, in so doing, they have acquired the right to occupy It. Because there is no mandatory collective registration of land, SD is aware only of registered parcels which have been entered into the Land Registry Files, or "Livre Foncier'. There is no Inventory of registered parcels by geographical units. It is therefore difficult to gain a rapid overview In the field of what is illegal and what is not. Private state land is therefore highly vulnerable to squatting. As a result, SD has found it cannot cope with the problem of illegal occupation of private state land. 7.13 In part because formal private land markets are not very active, Malagasy families turn to two other sources to get land. The first source is the Central Government. Since public land reserves are quite small (only about 1,000 ha in Greater Antananarivo for an annual incresse of 50,000 people), the demand is essentially impossible to satisfy by this means. Moreover, Government land is usually sold below market price, so that the system allocates land improperly, undercutting the private markets. Individuals are furthermore given great latitude in choosing their parcel, which leads to high administrative costs for the State. In addition to froehold subject to the site improvement rule, the basic land law provides far State grants of leases of up to 18 years, and long-term leases (baux emphytheotiques) of 18-49 years in length. )_01 However, the prospect of illegal occupation at very low cost Implies that few fmilies Choose any of the legal options. 7.14 The second source for families is the Fokontany. In keeping with the long Malagasy tradition of local self-reliance, the President of the Executive Committee of an urban Fokontany may grant permission for a family newly arrived in the city to settle on a plot of land. Whether this land is already privately owned, for example, may not always be the ruling factor. The deference paid to the Fokontany as inheritors of the fokonolona (village assemsbly) tradition, and to its president, may weigh as much or more in the balance as the legal status of the land. 300 Law 60.004 of February 15, 1960. - 84 - B. Options for Improvement Obiec,;ives 7.15 Future actions in urban land management should be taken in support of three main objectives. First, centralization should be reduced and greater involvement of local sovernments.promoted in the administrat'on and management of both private and state lands. Second, intermediate procedures toward the legalization of traditional land practices should be identified. Examples of such procedures include delivery of occupancy certificates that would confirm the right of residents to remain where they are. Achievement of this objective, in turn, will require deeper knowledge of informal land transactions at the household and neighborhood levels. The third objective is to introduce incentives to facilitate the adoption of these intermediate procedures by both land occupants and local administrators. 7.16 Under the first Urban Project (Cr. 1497-MAG), the Urban Land Legislation Study has begun this process by identifying the major constraints affecting land management practices in urban areas 31/. The report recommends that several actions be taken to revise the existing urban land and urban planning legislation. The most important of these recommendations have to do with (a) improving the procedures for development .(subdivision) of.urban land; and (b).-increasing the supply of developable urban land. Improvina Land Development Procedures 7.17 Simplification of Planning Documents. Existing planning documents are known to be outdated and unresponsive to local needs. They should be replaced with two types of plans. The first, a "Plan des zones,a would be a simplified map of land uses in broad categories such as "residential,n 'commercial", "public right of way", etc. This use of broad categories, and the implication that boundaries between land uses will not always be precise, distinguishes the "Plan des zones" from a conventional zoning map. The second, % "Plan des zones inconstructibles et des alignements," would define the areas of the city where houses should not be built because services would be too expensive. These new plans would also serve as monitoring tools for existing occupation patterns in the city, as well as a basis for determining the exact boundaries of State lands. 31/ URBAPLANIJURECOISOMEAH, LUgislatlon domaniale et fonciere en zones urbaines, Final Report, 1989. - 85 - 7.18 Simplification of Subdivision Procedures. Procedures for the dividing of tracts of land into individual plots are highly complex and centralized. An effective alternative to the current system would be the development of serviced plots (parcellaires d'accueil). Serviced plots may have only basic road access and some drainage at the beginning. Land reserves are set aside for future improvements in services, which are introduced as the incowm of the residents improves. This system has been encouraged by DAUH, and programs Initiated by the Fivondronana of Antsirab; and Mahajanga are ready for implementation. In principle, public authorities can readily expropriate private land and use it for purposes such as this. Appropriate legislation exists for expropriation, but alongside it is found other legislation which may contradict its provisions, 7.19 Simplification of Building Requirements for Low-Income Housing. In poor neighborhoods, where the regular system of building permits does not operate, a simpler system should be introduced. It would involve only two steps. The procedure vould start with a request for a Settlement Permit (Permis d'implantation), which the applicant would forward to the local government concerned. The request would include the size and location of the building, plus, if available, evidence of occupancy rights or confirmation from the legal owner of authorization to build. One year after the Settlement Permit is granted, the construction would be submitted to a building inspection which would verify that site improvement (mise en valeur) has occurred. If so, an occupancy certificate (certificat de mise en valeur, or permis d'habiter) would be granted. This procedure would apply only to low-income housing; building permits for high-standard housing would be delivered in the normal manner accoreing to Articles 104 and 105 of the Urban Planning Code. Renting rather thar home ownership is the only viable option for a great many Malagasy urban families (see Chapter VIII). The Settlement Permit would not necessarily be linked with ownership, but is likely instead to be granted in conjunction with the payment of rent for the property. 7.20 Establishment of a Land Reaistry and Data lank. The Settlement Permit would provide information on new construction and would consequently help with updating of the land registry files (fichier foncier) at the local government level. This data bank would include information on the ownership and tenure status of the land, its location, and the characteristics of the owner or occupant. Linkages should also be established with the ongoing inventories and field surveys of the Department of Taxation, so that the data bank can capitalize on information collected during tax assessments (see also paras. 6.16 - 6.19). 7.21 Incentives to Local Governments. Since this system will take some time to become fully operational, complementary ways can be found to deal with illegal occupation. One approach would be to develop contractual agreements between DP and local governments, in which some State lands would be made available to the city for low-income housing in exchange for protection of other State lands, especially the private domain of the State, against squatting. - 86 - Incr.asina the 8 i1l of Develoiable Land 7.22 Tho direct *pproaeh outlined above, in which the evail.bility of State lands for low-income housing Is the subject of negotiation between the Central Government (DP in this case) and individual local goornments, is feasible If suitable procedures are put in place. It also has the advantage of allowing solutions to bo tailored to the special circumstances of each city or toan. However, it Is also intensive of both staff and time. For several years, the Idea that econoamies of scale in this process could be realized through the cre&tion of a land development agency (Agence fonciere urbaine, AFU) has been debated among Malagasy urban specialists. There are *c-nd arguments against the creation of such an agency, not least that it may iLuolve large budget outlays at a time when economic policy calls for the gove 'ument sector to be changing from direct management to coordination. It is n \ zrtheless useful to assess the merits and demerits of such an option, so tbat the question of whether or not to move forward with such an agency can be made rapidly. 7.23 Instead of allocating land on an Individual basis, the State would initiate the allocation of large parcele and delegate the management of the process to an AFU. The AFU would receive from the DP parcels belonging to the private domain of the State. These parcels would be registered in the name of the AFU. The agency would establish simplified urban plans (see pars. 7.17) to be approved by the local government and the Ministry of Public Works. This plan would show the future location of infrastructure, public buildings, and the road network. AFU would sell the plots to benefiolaries whb oeet the eligibility criteria previously agreed upon. After purchaae of the plot, the individual vould request registration of the plot in the Land Registry (fichier foncier). Long term leases could he developed to reach the lower income groups. 7.24 In secondary and smaller towns these procedures could be further simplified. The State would cede some of its private domain to local authorities in order to facilitate low-income housing under conditions of a long-term lease with option to buy. The Fivondronana or Firaisana would deliver occupancy certificates which would involve payment of a monthly rent. The certificates would be changed into full ownership titles upon expiration of the lease. Local governments would also need to activate the planning process to prevent the encroachment of private development on land allocated for public use. c. Implementation Issues 7.25 Given the traditional reluctance of the Central Government to reform urban land legislation, an incremental approach is recommended. This approach Is the only one that has a good chance of success because (a) the - 87 - reforms Involve strengthening of the capacity cf local authorities, which will take time to reallseg (b) some Malagasy officiials do not view urban land Issues separately from rural ones, -which may lead to a protracted national debate on legislative reform; and (c) results will be obtained more quickly if the process begins with elements on which a political consensus has already been reached. It should be clear, however, that Incremental does not sean piecemeal. The interrelated nature of the changes proposed is such that the success of one component depends on the Implementation of another. For example, procedures for building permit approvals depend on evidence of land tenure provided by the registration system. Therefore, g clear vision of the results to be obtained from all measures must accompany the phased introduction of any one measure. 7.26 This vision if the future system of urban land management Includes two principal roles for the Central Government, neither of which it is performing now. Current legislation is base on direct control of land uses by the Government. The reality Is virtually the opposite of direct control. One element of the future role of the central authorities is therefore to guide and facilitate urban land management rather than to control it. To do this the Government would make land available to development agencies or associations, public or private, for all purposes. The proposed creation of the AFU leads in this direction. It would also promote the adoption of the simplified urban land and urban planning instruments described above. The second critical role of the Central Government is actively to promote a partnership with local government in the management of urban land. Initially, DP and DAUH may still handle the delivery of occupancy permits, approval of urban plans, and land registration. It is recommended, however, that in Antananarivo and in the Faritany capitals this resnonsibility be formally delegated ta DP and DAUH (maltres d'ouvrage delegues) by the local authorities (maitras d'ouvrage). Signature of a protocol formalizing this relationship vould not change in any way the day-to-day activities of DP and DAUH. It would, however, strengthen the participation of local authorities in the process by holding them accountable in law, as they already are in fact, for how their patrimony is used. DP and DAUH resources would be at the service of local governments and people in law, as they already are in fact. 7.27 In practical terms, the vision of Government as facilitator and guide in urban land management should be realised in phases, and is likely to take considerable time. Three phases, lasting 4-5 years each, are recommended. They are linked with the 3 phases of the effort, about to begin under the Environmental Action Plan with assistance from the World Bank and other donors, to clarify the status of land parcels totalling some 55 km2 in Antananarivo, the 5 other Faritany capitals, and Antsirabe. 7.28 The first phase would begin with an opinion from the Urban Development Consultative Committee (CTNDU) on the directions that reforms In urban land management should take, and on the sequence of actions to be carried out in support of these reforms. For this to occur effectively, DP will need to become a full member of the CTNDU and to be represented in all its deliberations. An inventory of private and private State lands is also - 88 - critical for effective urban land management. A rapid inventory of lands belonging to urban CDs will allow these local authorities to learn the extent of their holdings. CDs should carry out this activity In collaboration with DP. 7.29 Actions in the first phase would include a general cleaning of existing texts, mainly to eliminate inconsistencies and duplicatior.. At the same time, drafts of the settlement permit, zone plan, and other instruments would be prepared for discussion. The legislation would not iritially be made mandatory in all cities and towns. Instead, these texts would authorile local governments to adopt the measures if they wished, and to call upon Central Government resources -.or their realization. In this manner the legislation can be tried in selected towns on a pilot basis, and later generalized if the situation warrants. 7.30. Another critical step in this phase is the strengthening of working relations between DP, the Department of Taxation, DAUH, and local governments, possibly as part of the technical mandate of the CTNDU. During this phase the cadastre program would begin the process of preparation of the physical cadastre, so that the boundaries of private State and private land would ultimately be known. 7.31 In the second ghase, the new syctem would be introduced in more cities and towns that sho'f commitment and interest, and decisions would be reached on the allocation of responsibility between DP and the local governments for land registry 32/. The cadastre program of the Environmental Action Plan will simultaneously be entering its most critical stage, that of the local adjudication of who precisely is the owner or occupier of land parcels. Contrary to the early estimates of six months, this process is likely to be spread over several years and require much staff time, negotiating skills, and patience to complete. Urban programs should themselves contribute resources to assist in this tnsk. The reforms in urban land management would complement this effort with field surveys, to determine the approximate size of all the household plots located within the larger tracts of private State land whose boundaries were earlier defined. In the case of Antananarivo, responsibility for land registration, settlement permits, and urban plans would be transferred to the city administration. 321 These decisions are likely to involve shared responsibilities, in which DP through the 'Services provinciaux du cadastre' maintains and updates the cadastre for private and private State lands, while the Fivondronana or Firaisana maintain the registry of properties subject to settlement permits. Much variation can occur, however, as c" tes will differ in their ability to manage the registries. For th .St chance oP success, a program of institutional strengthening through equipment and training should be put in place so that urban land managers can become more technically proficient. - 89 - 7.32 The third phase would consist of further transfer of responsibility for land and urban planning to Fivondronana or Firaisana of major urban centers1 still with guidance froir DP and other central agencies. The prospect of constituting land reserves fo.- future urban development should be considered only at this time. Land reserves can be an expensive cholce, and should be considered only after much experience is gained with other, less costly options. The cadastre program during this time vwuld proceed with delivery of land titles to those eligible, while settlement permits would continue to be delivered to others. 7.33 As discussed in Chapter VI, land taxation and its relation to land registration and cadastral Improvements would need to be a focus of all three phases. The initial attention to improvements in the property tax base and in collections would be followed in later phases by a more basic reflection on the structure of the property tax and its incidence. ihis reflection, however, needs to begin now, so that any fundamental reforms decided upon are not unduly delayed. - 90 - CHAPTER VIII HOUSING AND HOUSING FINANCE A. The Foundations of Housing Policies Urbanization. Public Investment, and Housing Finance 8.1 The challenges confronting Madagascar's economy in the 1980s have contributed to a marked deterioration in urban housing conditions. The decline in real household income and in the capacity to save have resulted in overcrowding in poorly serviced neighborhoods of Antananarivo, growth of squatter settlements in secondary cities, a decline in the general quality of housing, and a fall in the supply of new dwellings produced by households and small investors. The housing market has stabilized at a level too low to absorb the growth of the urban population (see also Annex 8.1). According to recent experience of SEIMAD, one of the main shortcomings of current housing production is its unfavorable ratio of quality to price. The housing constructed is too often of poor quality and at the same time too expensive for families it is designed for. 8.2 Reforms have meanwhile been introduced to establish the basis for a resumption of real per capita income growth. New public expenditure and social programs were prepared and trade liberalized, while financial sector reforms have aimed at reducing the traditionally large bank credit flows to public enterprises so that the private sector could improve its access to credit. A critical national objective henceforth is to sustain the real per capita income growth that began in 1988, while holding down inflation and the deterioration of the external accounts. This implies that in the short term little relief for housing can be expected from public expenditure. Additional resources resulting from liberalization will be allocated very selectively, as part of difficult tradeoffs among sectors and projects. 8.3 There are nevertheless good grounds for promoting investment in the housing sector, both to alleviate the social impacts of the liberalization program and to make a positive contribution to overall growth. Given the prevailing patterns of production, an increase in the output of the construction sector could be achieved without substantial infusions of capital. If the production of low-cost housing is encouraged, the effect on tb-e trade balance would be manageable and the materials and labor required would stimulate local employment. In light of the objective of tight management of Internal demand, support for the housing of employees - 91 - of firms or public enterprises could be an effective way to distribute purchasing power to wage earners. 8.4 Earlier sections have stressed the importance of increased investment by central and local governments in urban infrastructure (see especially Chapter IV). In view of the volume of resources required to have any significant impact, most of the public funds for the urban sector must be allocated to such investments. Necessarily, therefore, financing the growth of the supply of dwellings will depend critically on the savings and investment behavior of households and firms. Once an affordable supply exists, increases in expenditure to improve housing conditions are decisions for households and firms to take. The role of the State and of local governments would change. In partnership, they would primarily act to (a) stimulate the production of serviced lots; and (b) take appropriate fiscal measures to promote private investment in the rental market. Other desirable steps would be to establish mechanisms to mobilize savings and credit for housing, to publicize these measures widely, and to follow up with other actions (principally regarding the production of building sites and savings-for-housing instruments) that would stimulate private initiative and gradually dissipate expectations of intervention by the 'all-providing State' (Etat providence). These actions would instead foster a climate of growth expectations, that is essential if the urban sector is to play its role in national growth and the desired changes are to be launched rapidly. In summary, the State would no longer spend public funds to construct housing. Its role would be confined to encouragement of individual and group efforts to build housing, and to facilitation of access to buildable land and to housing credits. 8.5 Following an analysis of savings and investment in urban housing, the remaining sections of this chapter describe instruments and programs for the financing of housing that are consistent with this approach. Among the most important topics in this regard are the linkages between housing finance and the banking system, the role of mutual savings and savings-for-housing programs, and the chances of improving access to rental housing. A concluding section outlines the elements of the recommended program. A sua-ary of features of the urban housing market is found in Annex 8.1. Savings and Investment in Housing 8.6 Financial programs based on improved mobilization of private savings are essential for urban housing to escape its current stagnation. This is a major condition to be realized if urban investments are to grow on a sustained basis, without worsening budget deficits, debt, or inflation. Discussions, studies and seminars organized over the past decade have stressed this point. At present, the need to finance housing through private savings rather than public funds has been fully integrated into the sectoral and national policies of the Government. - 92 - 8.7 Household savings depend on the level of income and also on its prospect for growth over time. At a given income level, the tendency to save for housing rather than to consume will depend mainly on political and fiscal incentives, and on the availability of buildable land and credit. As these elements of urban development policy are put in place, it can be expected that saving for housing, at the top of the list of a household's assets, will respond positively. Other critical factors in this growth of savings are the structure of income distribution and the extent of participation by middle- and higher-income groups, who have not yet been systematically appealed to in this respect. 8.8 Based on the rural-urban linkages study conducted by Groupe Huit/AURA in 1988-89, about 4% (FMG 16 billion) of the FMG 447 billion of annual gross household income (before taxes and contributions, and including non-monetary income and transfers) represents investment in construction. Slightly more than 8% (FMG 35 billion) is saved, with FMG 31 billion of the latter in monetary form. This capacity is very unevenly distributed: the 3% of households (expatriates, senior officials, and local businessmen) who account for 422 of total incomes are responsible for 902 of savings and 302 of real estate investment. 8.9 Savings and investment by the approximately 165,000 other urban households in the Antananarivo region are relatively low. These households save FMG 7 billion (FMG 4 billion in non-monetary form) and invest FMG 11 billion in property (FMG 3.5 billion in the form of work on self-built housing). The savings rate of households earning a portion of their income from agriculture, or from artisan trades and traditional services, is somewhat higher (6-9S of Income) than for company wage earners or government employees (22 of income), even though the latter's income levels are twice as high. Wage earners and government employees, on the other hand, account for FMG 6.3 billion and 2.3 billion respectively of construction, or 55% of total real estate expenditure. Even at low incomes, therefore, people do manage to save, and the potential to improve savings exists at all income levels. 8.10 Investment by households of FMG 16 billion represents an output of 3,200 dwellings, including 200 high-standard units (costing FMG 25 million each on average, with financing coming where necessary from bank loans), and 3,000 dwellings financed by the households themselves through informal channels. All socioeconomic groups derive income from rentals (3- 41 of total income on average). Currently, about half of the PMG 11 billion of annual rental income appears to be reinvested in construction, resulting in an annual output of 1,500 units from this source. Access to Property and Rental Housing 8.11 The knowledge of who can invest, In what forms and for what type of housing, determines the strategies to be adopted for mobilization of savings. These strategies will vary according to socioeconomic groups and to the financing mechanisms and type of construction to be promoted. New - 93 - construction is more readily affordable to middle- and higher-income households, who also tend to have lived the longest time in the city and to already have housing. Savings mobilization and housing loans should be directed as a first priority to these groups. 8.12 Within new construction, an important distinction is made between housing built for personal use and that built for rental. Rental housing can be undertaken by individual investors, generally owners already, building a modest extension of their present house or a moderate-cost dwelling on another plot. These houses are generally offered for rent to low- or middle-income families. It may also be carried out by institutional investors, consisting of firms or developers who can assemble more capital and whose housing is offered at higher rents to better-off families. Therefore, housing finance programs targeted initially at middle brackets and even higher are still compatible with social welfare, since lower-income groups also benefit through the rental market. 8.13 In contrast with past programs, any future promotion of home ownership should not overlook this important role of the rental market in the shelter of the poorest households, as well as in mobilizing savings. Rentals account for 502-602 of the urban housing market, and up to 802 for the lowest income groups. Renting is prevalent among all groups, not merely for lack of alternatives, but also because it affords better access to urban services, infrastructure, and transportation than owner-occupied housing which may be relegated to the periphery because of the lack of buildable land. 8.14 The rental market therefore provides the means by which the majority of urban households can improve their housing. 33/ The upward movement will occur rapidly to the extent new dwellings are built on sites that can be developed immediately, and as neighborhoods lacking proper infrastructure are upgraded. As average income earners move out of poorly maintained or overcrowded rental housing, the supply available to those with very low incomes will increase. The main incentive to Improve these dwellings will come from their owners' desire to increase the value of their assets as the urban environment is improved. Since few households with above-average incomes now enjoy good housing, those looking to move generally state that they are ready to pay higher rents for better accommodation. 8.15 Small-scale property owners, who sometimes own several houses of low value, are found in all population groups but especially among artisans, merchants, and government employees. They finance their investments outside the banking system, mainly from rents or the reinvestment of the proceeds of other assets. This kind of investor, saving to buy property, can be 331 A minority of hot 3holds with very low incomes still must improvise construction of a makeshift dwelling on land with uncertain title or on public lands. As progress is made in definition of the boundaries of such land, however, these prospects are bound to diminish. - 94 - encouraged by the availability of bank loans that add to his self-financing and hence accelerate the investment. Financial institutions should be able to attract the deposits of this group, which offers adequate guarantees. This part of the market is not large. Savings are presently low, and when accumulated t"nd to be devoted to supplies and other essentials for business activity. Nevertheless, extension to these groups of housing loan and savings programs can have an impact on their investment in housing, through the use of guarantee funds and specific forms of lending (e.g., for a dwelling used simultaneously as house and business). 8.16 A different problem is faced at the high end of the scale. The housing market for entrepreneurs, senior government officials, and expatriates is currently recovering, but will quickly reach the saturation point. As rental earnings grow, owners will look for profitable placements for their funds. They could be encouraged to turn to dwellings of medium quality, for example through tax advantages to which these investors are more sensitive than the small-scale owT:ers. 8.17 The desire for home ownership is strongest among the groups earning between the average and three times the average income. Lacking a savings channel in the financial sector, some of these households, in traditional occupations and among company wage earners in particular, launch their own construction through the informal sector. Dwellings built in this manner cost on average FMG 3.5 million (US$2,250), but the range is very wide, from FMG 1.5 million up to FMG 10 million (US$6,400). Families accumulate savings slowly, rely on help from others, occasionally take out loans secured by pledges, and build in phases. This process could be greatly improved by enhancing the return on savings and reducing the time required to obtain funds for construction. Such measures would appeal to these savers, who represent the largest potential group of customers for the financial institutions. 8.18 On balance, relatively favorable conditions therefore exist for the development of savings, in view of the inadequate housing conditions and low outlays on rent. Precoaditions for this development mainly include increasing the supply of buildable land, but also involve discouraging families from continuing to wait for low-rent public housing, and establishing suitable savings programs in the financial sector. B. Savings Mobilization and the Financing of Housina 8.19 Over the past decade, several plans for mobilizing savings and credit have been prepared, but have not been carried out. Although it is not necessary to start from scratch, the lack of implementation demonstrates that mobilization of private savings cannot be decreed. Savings mechanisms and institutions must be based on the needs, capacities, and behavior of all groups of the populatiou. Realistically, one should think not in terms of a single financing system but of a number of systems. This is especially the - 95 - case as the institutions taking in deposits may be different from those ensuring their transformation and distribution in the form of credits for housing. The quality of service and the ability to expand steadily while maintaining financial balance, rather than the number of dwellings financed, should be the goals of housing finance programs In the near term. Mutual Associations and Cooperatives 8.20 There is a real need in urban areas to develop intermediaries between the savings of individuals or groups on the one hand, and the banking system on the other. The family and the Fokontany are only two examples of Malagasy social solidarity. In government departments and private firms some organizations of a cooperative nature provide mutual assistance. 341 Except for certain modern sector firms, however, this social solidarity does not carry over into economic and financial matters. 8.21 These groups nevertheless constitute a basis for the organization of mutual savings programs, including cooperatives where appropriate, for the purpose of developing savings and small loans for housing or other uses. Many households, even those with average incomes, find it difficult to move quickly from informal supporG networks to a bank loan. This is a major socio-cultural change, for which mutual savings societies may provide an effective transition. With such groups, individuals retain a high degree of control over their funds. For example, several different deposit and tiithdrawal procedures can be developed within a single group (on demand, at agreed intervals, etc.). Conditions of loans given by the groups can vary according to the maturity and the frequency of payment of group members. Cultural problems posed by written contracts for financial obligations may be more readily overcome if others share the obligation. Savings, moreover, can remain liquid and accessible for current expenditure, with the added security that a loan can be obtained in unforeseen circumstances or for a limited investment. 8.22 In return for the advantages of flexible deposit and withdrawal privileges and of borrowings limited to the group, loans made by mutual savings systems cannot, without running substantial risks, exceed the volume of deposits received from members. 8.23 Mutual savings societies must offer market rates on savings in order to be competitive, and also for the return on savings to their members to keep ahead of increases in the cost of building materials. The cost of loans can nonetheless be kept reasonable by holding down operating and risk management costs. In cooperatives whose aim is to finance construction, difficulties are often encountered with loans grouped together, if construction or bridge loans are mixed with loans for house purchase. To 341 Examples are the Mutuelle des enseignants, Mutuelle des chefs de service de l'Administration, and the Mutuelle SEIMAD. - 96 _ succeed, these systems require cooperative members to have similar incomes and a strong sense of solidarity or, failing that, group pressure. 8.24 An important role of groups of this type is to develop a learning process about savings, which can subsequently lead to closer association with the banking system. Individuals with few alternatives and without much financial weight become collectively in a better position to negotiate the conditions under which their savings will be remunerated, and loans from the banking system granted. This benefit stems largely from the collective guarantee that can be offered to a lender by an individual borrower within a mutual association or cooperative, particularly for housing. 8.25 Actions to promote savings societies can be taken at two levels. First, the Central Government should define model statutes and other elements of the legal framework for efficient operation, as well as guidelines for the collection and use of funds and for the protection of association members. If members so desire, a private service agency such as the recently created SP MIEL can be hired to help with establishment of the association (see also Annex 8.3). Several non-profit foreign organizations have similar expertise and can also be called upon. It is important that assistance of this type be voluntarily entered into, however, and not required. 8.26 Agencies providing legal and management services should not also be the ones providing the capital. In this regard, large injections of credit to the savings associations, which could exceed the savings capability of the members, should be avoided. 8.27 Another type of assistance is a second-tier guarantee of the societies against their member-borrowers' inability to repay loans. This guar&ntee would not affect the need to recover the loans, but would if necessary bolster the funds of the association. The Central Government or local authorities could provide this guarantee and also make capital contributions to mutual guarantee funds vis-a-vis lenders in the banking system. 351 It is imperative, however, that guarantee funds be managed efficiently and economically. Finally, an additional Incentive could take the form of tax relief on the interest from placements and on loans made by the savings societies. The Banking System and Housing Finance 8.28 The development of new instruments for savings and housing loans can be expected to promote both the ongoing sector reforms and a better distribution of loanable funds. These instruments should originate in the 351 A suggested formula is to split the second-tier guarantee into equal parts, with the Central Government, the local authority, and the agency in question each taking one-third. - 97 - banking system, whose iim Ing situation, resulting from reforms initiated in 1988, will enable it to Flay a steadily larger role. 8.29 The banking system, in serious difficulty until quite recently, is now being opened to greater competition. In addition to the Central Bank, the centerpiece of the system has been three State-owned commercial banks: BNI (industry), BTM (agriculture), and BPY (commerce). Although specialized in these sectors according to 1977 legislation, these banks in practice have played only a minor role in their designated sphere of action, and now have identically diverse portfolios. This arrangement has been supplemented since October 1989 by a fourth full-service bank, BMOI (Banque malgache de l'Ocean Indien), an affiliate of the Banque Nationale de Paris (BNP), with a minority Malagasy participation. BFV is also expected to have a minority private sector holding. 8.30 Provisions of the banking reform were defined in the 1988 Banking Law. 36/ The State banks are to become corporations (societes anonymes) open to private capital. Foreign-owned banks whose activities are consistent with national development strategies may become established in Madagascar. Also under this Ordinance and subject to Government agreement, specialized financial institutions may be established to improve the flow of funds to specific sectors, including housing. For the State banks the provisions made for writeoffs of virtually non-recoverable debts (accounting for 30-60% of their portfolios), the ongoing replenishment of their capital, and the enhanced repayment capacity of the public sector as a consequence of debt relief, should release new funds for investtent. The Central Bank is currently preparing regulations that would promote an increased supply of medium- and long-term credit (which at the end of 1988 accounted for only 182 of bank credit), andtfacilitate.the establishment-of new institutions. 8.31 Thus far, the reforms have accomplished the major step of convincing the banking and financial community that a much improved enabling environment is being put in place. The setting for greater emphasis on the financing of house construction is therefore more favorable. Transformation of this improved setting into practice, however, will require a good deal more effort. First, through the ongoing financial restructuring the State banks must improve the quality of their portfolios it order to provide good value from their large staffs and network of some one hundred branches. This effort should be pursued vigorously. 8.32 The improvement in the financial situation of banks will still not, however, be sufficient for an expansion of housing finance. The very low involvement in this sector stems not only from the difficulty of making long-term loans at low interest, but also from lack of concern and attention on the part of banks to household financial services. In the past it has Indeed been more costly and complicated to cater to small borrowers than to make large loans to industry or short-term, high-margin trade credits. There is currently no savings Instrument, other than term deposits, which earns a good return and is designed to appeal to large depositors, whether 361 Ordinance 88-005 of April 18, 1988. - 98 - firms or individuals. Conditions for housing loans are quite restrictive, as they have needed to be because of the very limited access of banks to the rediscount facility of the Central Bank. The maximum loan maturity for housing is still 5 years, with a maximum loan to value ratio of 70X, loan limit of FMG 35,000 per m2 of livable space, and loan ceiling of FMG 5.6 million (US$3,600). Nominal interest rates of 202 and more discourage borrowing over the long term. Access to the rediscount window is limited by strictly applied ceilings on loans outstanding set by the Central Bank. 8.33 The ternm transformation capacity of banks is likely to remain limited in the foreseeable future, and competition from sectors other than housing will undoubtedly grow even stronger. Housing loans will therefore need a permanent, specialized source of funds. This essential base can be obtained from household savings, and can amount to some 25-301 of the total funds able to be mobilized for housaing. The growth in household savings should be promoted not only by the commercial banks and specialized affiliates they may establish for the purpose, but also by the national Lnsurance companies (ARO and NY HAVANA) and national savings funds (Caisse nationale de prevoyance sociale [CNAPS], and Caisse nationale d'epargne), which also mobilize the savings of households and firms. 8.34 Housing finance, in summary, represents a major opportunity for the banking system. In overall terms the risk is small in view of housaing's fragmented nature and the guarantees available. Profit margins would be sm-all to begin with, but the effective demand at the upper- and miedle- income levels should ensure a steady return. More recourse to computers in the processing of loan accounts, plus greater control of overheads, should contrioute to this return on the cost side. A special effort in training and marketing of new loan and.savings products will also be required. Savinas-for-Hous ino Programs 8.35 Among the wide array of savings-for-housing programs, priority should be given to those based on a contract between the depositor and the financlal institution, so that the terms of the loan can be adapted to local conditions. In such schemes the saver undertakes to deposit a given amount at an agreed frequency. In return, the bank guarantees to provide a housing loan of an amount and at terms that would be a function of the deposits made, borrowing and lending rates, and pre-established profit margins 37t. This system meets a number of needs. For the saver, the main attraction is the certainty of the loan. For the lender there is the prospect of reasonable control of commitments by choosing appropriate lending terms and adjusting them over time in accordance with cash flow positions. The initial savings phase reduces the size of the loan required 37| BMOI in March 1990 launched a scheme of this type, which appears to have enjoyed a good measure of initial success. Contractual savings accounts for housing are paid 82 interest over the required 5-year savings period, and other savings accounts are remunerated at 5-7.5X. - 99 - and enables both parties to test their capacity for sustained financial effort. Annex 8.2 summarizes the main features of contractual savings programs and provides some illustrative examples. Annex 8.3 describes and evaluates a new entity ("MIEL") created to address these concerns. 8.36 The interest rates offered on savers' deposits should not penalize savers. A saver who does not borrow must be able to preserve the real value of his capital. At the same time, the certainty of a loan at the end of the saving period enables the rate on deposits to be kept lowar than for other financial instruments of the same maturity. 38/ The Interest rates on loans made are linked to the borrowing rates (i.e., the rates offered on deposits), plus a lender's margin. In such a system the marg"i would need to be considerably reduced from the ones currently prevailing. To launch the system effectively, selected fiscal advantages (e.g., exemption from the TUT and deduction of amounts saved from taxable Income) could be granted by the authorities. 8.37 As one example, it appears feasible In the current financial context that a savings deposit be paid 921, with loans then carrying Interest of between 12% and 142. Competition among lenders would be a critical factor in reducing margins. If lenders begin in this situation with a limited transformation, the housing loan would be for about the same amount as the accumulsated savinigs. Monthly payments on the loan would be close In amount to the Interest received from savings, rasulting in minimal disruption to household budgets. The repayment period for the loan, It should be noted, would be about twice as long as the savings period. 8.38 Systems of this kind are designed to appeal mainly to wage earners with steady, above-average incomes. They in no way exclude, however, either households in the traditional sector or high-income households, who can make substantial deposits with a view to purchasing property. To prevent large depositors from taking most of the loan funds, loan ceilings can be adopted, with a premium interest rate applied to any savings above the ceiling. Other Resources of the Banking System 8.39 Savings programs for housing can succeed only ln a sound financial sector environment. A feature of such an environment ±s that interest rates and the volume of loans are not set by policy, but are determined by market forces within generally agreed principles of prudent management and money supply limits. At present, several economic trends are converging to give rise to very high nominal and real interest rates for lending (7-10 percentage points in some cases). There is good potential 3u1 More generally, the introduction of savings-for-housing programs, with typically lower interest on savings because of the guaranteed loan, does not rule out the development In parallel of other, non- contractual savings instruments offering higher ret.urns. - 100 - for a reduction In these rates over time. Margins, as discussed earlier, can also be gradually reduced by modernizing procedures, computerization, by reducing the risks of insolvency, and especially by greater competition. 8.40 Other measures can be taken to reduce the cost of borrowing. Exemption from the TUT for loans can be considered, but will require special measures to avoid this advantage being "recovered" by lenders. An alternative would be to reduce or eliminate the various charges (frais annexes) associated with the loan. Registration duties alone account for 142 of the cost of many loans, with all charges together as much as 20S. 8.41 Beyond the basic resources provided by contractual savings and other programs, there is the larger problem of the supply of medium- and long-term loanable funds required for lending on a scale commensurate with the potential demand. Over the medium term the banks or their specialized affiliates will benefit from increases in their capital, but this is not likely to be enough. Consideration could be given to the introduction of a "Develo2ment Account," as one element of the secondary market. The Development Account would refinance loans by financial institutions (other than mutual savings groups) with terms greater than five years, as close as possible to market conditions, and would be managed by the Central Bank. The account could be provisioned by issues of bonds and other obligations to the insurance companies, CNAPS, and the public, as well as by any surplus funds from banks' savings-for-housing operations. The terms and amount of refinanceable loans, and the relevant interest rates, would be mutually agreed by the bank and the Government. A State guarantee and possibly also tax exemption should enable the obligations to be issued at a competitive cost. A balance would need to be struck so that the issue of these obligations gives some Impetus to the housing sector withoutAInfringing excessively on Treasury requirements for funds for other uses. 8.42 Measures should also be taken to improve and diversify the types of loans offered. The introduction of short-term bridge loans to developers would fill a gap In the current loan structure. These programs wotild be kept distinct from conventional mortgages, although they could both be offered by the same bank. Conventional loans should at the same tima be expanded so that loans can be obtained for purchase of housing, construction, improvements, adding rental space, commercial and artisan activity, and purchase of materials. Construction carried out by artisans should be accepted (uj to a limit), to the extent this does not interfere with the required loan guarantees. 8.43 Banks may be concerned, even anxious initially, about loan guarantees. One reason that few housing loans have been made is that as discussed In Chapter VII, the property registration system has not been producing full property titles, which banks have traditionally required as collateral. A lien on the property is not, in fact, a guarantee of great value because of the difficulties of enforcement. It nevertheless serves notice that the bank at all times retains the right to dispose of the property in order to collect on its loan. The same purpose can be served by the proposed settlement permit as by a full title. As experience in other countries (e.g., Rwanda) has shown occupancy rights have value and can be - 101 - monetized by banks in payment of debt. If, in addition, other re'.able guarantees including mortgage insurance, mutual guarantees, and guarantee funds are introduced, there should be ample security for housing loans. A Housing Finance Institution? 8.44 Given the desire to expand access to housing loans, as well as the synergy from having many housing sector specialists in one place, it is natural to consider whether to create a specialized housing finance institution. On balance, the costs appear to outweigh the advantages. Considerable capital would need to be assembled for such an institution, and this capital must be compensated. It is in fact hard 'o imagine that a specialized agency, operating under market conditions, could do better than existing banks and their affiliates, especially as it would have far fewer branch locations. Any subsidies it requires to make loans more accessible would undoubtedly be captured by better-off families, able to use the funds more quickly. Experience elsewhere has also shown that the granting of loans can give rise to local political interference, that is generally more difficult to avoid than in a full-service banking system. 8.45 Some intermediate measures are possible, however. Under the current reforms existing banks, individually or in a group, may choose to develop housing loans through a specialized entity created with their own capital. In a second phase the entity could attract outside capital. Another arrangement, alluded to earlier, is for savings to be mobilized by existing bank branches while a new entity handles complementary resources (bonds and other. obligations,,Central Bank rediscount, bank lines of credit) as well as portfolio management. gmployee and Civil Sarvant Housing 8.46 The participation of firms in financing housing for their employees could be increased, especially as many have shown themselves anxious to solve this problem. Tax exemptions on earnings from amounts invested for this purpose, and enhanced returns on savings deposits for housing, are likely to encourage firms in this direction. As for the 5,000 civil servants housing units, one solution which should be explored is to sell the units to their legal occupants. Receipts from these sales could be placed in a fund for the construction of new units for either civil servants or the population at large. An array of alternatives should be drawn up, based on the agreed policy that the State will no longer directly invest in housing. - 102 - C. Sumamry of Recommendations 8.47 In a strategV or iolicv statement, the Government should clearly indicate its choices in the area of housing iolUcY. 8.48 The most important element of this strategy Is to imorove the pLoduction of serviced lots. The 'plateaux dthabitatO project assisted by EDF, and the urban projects of the Ministry of Public Works financed In part by FAC (France), are Important Initiatives in this respect (see also Chapter V). These programs would begin mainly in the suburbs of Antananarivo, the other five Faritany capitals, Antsirabe, and selected smaller towns. In the areas to be developed through public initiative, all types of investors should be encouraged to construct rental as well as owner-occupied housing. Speeding up the production of serviced lots Is the most effective method of Increasing tho number of potential savers and builders. 8.49 The encouragement of rental housing should consis- first of allowing rent increases to follow the market, and next of gr ating temporary tax exemptions for new rentals (with partial or total compensation paid to local governments for revenues foregone). Other recommended measures Include deduction from taxable income of a portion of rents received, and access to savings progrms for financing additional rental space. The oecupation of public rental housing by families who can pay higher rents (see Annex 8.1) should be phased out through regular increases in rents. A portion of the Increased rental receipts may be assigned, at least initially, to savings for housing accounts. 8.50 Promotion of serviced lots complements two other critical measures. The process of strenxthenina local government is enhanced through initiation of projects at the local level and delegation of project responsibility by local governments to public or private executing agencies (856EIM, for example) or to groups of residents. Second, families who manage to afford housing typically rely on informal construction networks. Partial tax exemptions to Informal or artisan activity may be justified as a spur to their development, especially as the tax revenues foregone vould likely be equalled or exceeded by the economic activity generated. The construetion of demonstration houses, with use of local materials, could also be carried out, with permits, authorizations, and other formalities held to a minimum. 8.51 A policy favoring hme ownershii would have two main features. It vould first assist in the establishment and operation of mutual or coonerative societies and associations for savings and credit. The Government can facilitate these associations by (a) helping to define their statutes and operating principles; (b) constituting guarantee funds; and (c) granting tax benefits on deposits and loans. Second, contractual savings iRograme would be introduced, on a competitive basis among institutions, together with an Information campaign on the aims and advantages of these programs. Encouragement would mainly take the form of reductions in registration duties, fiscal Incentives, exemption from the TUT on loans and - 103 - savings, and Incentives to improve older dwellings. Company profits earmarked for employee housing could be exempted from taxation, and bonuses could be paid on housing savings deposits opened by wage earners. 8.52 Central Bank reimlations should establish the guidelines, but not the details, for banks and other financial institutions to launch effective housing finance programs. For interest rates, savings periods, size of the loan, and other detailed features, maxiuim latitude should be left to the banks to compete with each other within the prescribed Central Bank guidelines. Any new legislation in this area should also be thoroughly discussed with the commercial banks themselves before it is put in final form. 8.53 The most critical financial sector measures described earlier include the revisions to conditions under which banks have access to the Central Bank rediscount facility for housing credits, and the possible creation of a Development Account open to foreign capital. 8.54 More attention to the production of serviced land and housing will furthermore benefit the market for construction and for building materials. Professionalism is not always in evidence in construction firms, making it difficult to determine their true competence. In this regard it would be useful to explore the possibility of launching a professional society that would promote greater discipline and transparency in the sector. A positive rating from this society, for example, could be a factor in judging the competence of the firm. In addition, building materials are frequently not available in the quantity and quality required. An expanding market for such materials Improves the prospects for standardized production on a larger scale of doors, windows, paneling, etc. The creation of production units In the PDUM project is an example to be continued and expanded. - 104 - CHAPTER IX URBAN TRANSPORT A. Introduction 9.1 Like other sectors of the urban economy, urban trarsport has been seriously affected by the economic stagnation which prevailed during much of the 1980s. The vicious cycle of low fares, poor roads, congested streets, and inadequate service has so reduced the supply of transport, especially in Antananarivo, that lew-income families are probably hurt fully as much or more than if fares had been increased. Vehicles are systematically overcrowded, further reducing the quality of servic- and increasing the cost of vehicle maintenance. 9.2 A major effort will be required to improve the effectiveness of public transport of passengers in Malagasy cities. For the present this effort should be concentrated in Antananarivo, where the magnitude of the problems is many times greater than that of other cities. The following sections examine the main issues that must be addressed to improve passenger transport in Antananarivo. 391 The final part of the chapter identifies six improvement measures that would result in passenger transport playing an effective role in rural-urban excnanges, employment, and growth. B. The Public Transport Crisis in Antananarivo Transport Operators 9.3 As is well known, 2 of the 3 main public bus companies (Societes d'etat) serving Antananarivo are in serious trouble. FIMA, created in 1978, 39/ In focusing on passenger transport in Antananarivo, this analysis addresses what appears to be the most serious problem of urban transport in the country. It therefore does not cover at least two issues which, although important, can be better addressed at a later stage: the transport of goods within and between urban areas, and passenger transport in the Faritany capitals. For the latter issue, one or two more years should elapse before the experience of thea newly created bus companies in the Faritany capitals should be evaluated. Such an evaluation will be quite useful, however, since there Is already concern about undercapitalization of some of these companies and about ridership which has not come up to projections. - 105 - is owned by the Government, together with banks and insurance companies. At present it is virtually at a stoppage of activity, with only about 20 vehicles out of 116 still in service. A consensus has been building that FIMA should be liquidated, but the right means of doing so, with 540 employees still on the payroll, has so far proved elusive. FIBATA, also established in 1978 and also State-owned, has now abandoned municipal service entirely in favor of suburban routes with small-capacity vehicles. Its financial situation has not improved, however. 9.4 The third and newest company, ANTAFITA, was created in 1982 by the Fivondronana of Antananarivo-Renivohitra. It is the only company presently offering service on large-capacity (80 passengers) buses. Its financial situation could appear better than that of the other 2 companies, until it is considered that no provision for depreciation of the vehicle stock is being made in its accounts. Its personnel costs and workforce have been increasing faster than output and turnover. 401 9.5 Confronted with chronic service failures on the part of the bus companies, users have grouped together into cooperatives to provide service, especially in suburban areas of the Antananarivo region. Cooperatives now account for about one-third of passenger trips in the city and two-thirds in the suburbs. Individual operators of taxi-brousse and mini-vans also make up a growing share, as bus service withers away. Being more flexible, these operators have adapted better to the demand and are essentially the only operators able to make a living in urban transport today. The supply of vehicles owned by cooperatives has recently increased, but the fleet remains old and drivers do not always respect the zones of operation assigned to them. Users 9.6 Even for the poor, the long distances between residence and work or school leave little choice for many families but to use bus transport. Between 1987 and 1988, despite a doubling of bus fares in the city fr*4 FMG 50 to FNG 100, passenger traffic on, for example, line 5 (Analakely - Andranomena) decreased by only 2.92. 411 For suburban transport, statistics seem to indicate a 15X increase in passenger traffic during the same period. When fares increase, riders generally cannot switch to other 40/ Financial indicators and other information on the bus companies have traditionally been quite difficult to obtain. An overview of the major issues confronting these companies is found in a report prepared by DEPCT (Direction des etudes, de la programmation, et de la coordination des transports) of the Transport Ministry, Transports collectifs de l'agjlomeration d'Antananarivo. Rapport d'etude prelmnaire sur un aroiet d'amelioration, April 1988. 41/ DEPCT, Transports collectifs de l'aatlomfration d'Antananarivot Plan de tranwport de la desserte du nord-ouest, Ministry of Transport, June 1988. - 106 - modes but have to absorb the Impact by not returning home for the midday meal or by contracting other expenditures. Fares 9.7 The fare structure varies according to whether the travel to urban (within the Fivondronana of Antananarivo-Renivohitra) or suburban. In Antananarivo the fare is a flat FMG 100 no matter what the distance traveled. There are no weekly ticket books or discounts for frequent users, and the present system does not encourage coordination or connection between lines. For suburban transport, fares are proportional to the distance traveled. 9.8 The Government has raised fares several times since 1975, but always with most attention to the very low purchasing power of users. As a result, for both urban and suburban transport the present fares do not cover operating expenses even before depreciation. Indeed, the faxes bring in so little revenue that, following the liberalization of imports beginning in 1987, the bus companies found they did not have enough cash to purchase badly needed spare parts and other items that were newly available. Moreover, enforcement of tariffs has been weak, and independent drivers offering more reliable service have operated at closer to market-clearing fares. Road Network 9.9 Because of the poor condition of the roads and streets, transport companies have abandoned some suburban and even urban lines. Central and local authorities have long recognized that the urban road network requires upgrading if it is no longer to constrain the growth and quality of urban transport. This has been recognized in the 'Charte Routire" 421, as well as In several studies carried out by the Traffic and Transport Bureau (Bureau de circulation et de transports, BCT) launched under the PDUM project (Cr. 1497-MAG). These studies document the condition of urban and suburban roads and the most urgent priorities for rehabilitation. Some progress has been made in Antananarivo with the upgrading of urban sections of the 'Route Nationale 4" totalling 23.5 km, and the installation of traffic lights at several major intersections. The next steps, however, are financial. Antananarivo municipal finances must bear more of the cost of such improvements than they are capable of at present (see Chapter VI). The Government should further ensure that part of the Road Fund resources are devoted to maintenance of urban roads. 9.10 If urban growth during the 19909 approaches the high urbanization scenario of Chapter IV, it will be extremely Important to have good road connections between cities and their hinterland. In particular, major arterials into the cities will have to be developed to take better account of the special nature of Malagasy road traffic. Man: pedestrians _42 Law 88-013 of October 25, 1988. - 107 - and carts use urban roads, and at the same time often play a significant role in their deterioration. Technical solutions should be found to accommodate this traffic safely, while protecting the part of the road designed for motor vehicles. Sector Orsanization and NHnagement 9.11 Until 1988-89, the basic law governing urban transport was a 1970 decree under which responsibility for urban and suburban transport was divided. 431 The Fivondronana of Antananarivo-Renivohitra was charged with the urban zone, while the suburban areas came under the jurisdiction of the Faritany. As the metropolitan area grew, this distinction gradually became an lmpediment to coordinated transport planning. 441 9.12 The issuance of a 1988 decree replacing the 1970 text, followed by an application measure in 1989, has brought greater clarity to the roles of central and local authorities in the management and organization of the sector, in two main respects. 45/ Urban and suburban zones have been merged into a single Metropolitan Zone. Second, a Technical Committee (Comite technique pour le transport urbain et suburbain, CTTUS) grouping ministries, transport companies, local authorities, and users is being established to advise on the preparation of transport plans and review operators' requests for traffic rights. This legislation also recognisea the reality of the market, by making it possible not only for bus companies sad cooperatives blat also for individual operators to receive a license to operate on a given route. If arplied flexibly (see below), the procedures and organization outlined in these two texts can provide a good enabling environment for urban transport In Antananarivo. C. Recovery Measures for Public Transport 9.13 A six-part program for the recovery of public passenger transport includes action to address all the above issues. The recommended measures are described more fully below. 431 Decree 70-074 of January 20, 1970. 44/ As one example, ANTAFITA, owned by the Fivondronana of Antananarivo- Renivohitra, had been unable to operate In the suburbs. It coulo not extend its service to suburban lines abandoned by FIMA. Cooperatives had become the only choice for the abandoned passengers. 451 Decroe 88-191 of May 10, 1988, followed by Arrete Interminist;riel (issued by the Ministries of Transport and Interior) 5115/89 of September 19, 1989. - 108 - 9.14 TransDort Companies. The future of FIBATA and FIMA has been debated for several years. Although prevailing views appear to be inclined toward liquidation of both companies, the consequences of this outcome for passenger traffic in Antananarivo, as well as for personnel of the companies, have not been fully identified. Some may also hold out hope for a restructuring. 9.15 The recently completed audits of FIBATA and FIMA address issues related to their financial and economic viability. It is recomended that the conclusions of the audits be widely circulated among ministries and agencies charged with making decisions on the future of these companies. 9.16 Sector Institutions. Provided they are flexibly applied, the key elements of the 1988-89 legislation should provide a solid basis for better performance in the sector. Urban transport according to this legislation will henceforth be developed according to transport plans, which would determine zones of transportation and routes and specify the main features of the demand and supply of urban transport in the area concerned. It should be recognized that the capacity to produce such plans, even within the Transport Ministry (MTMT), is still quite limited. As the capacity to prepare such plans is built up, it may be wise to focus scarce resources on the formulation of high-priority investment projects and their evaluat.-.on in the annual Public Expenditure Reviews (PERs). Worthwhile projects should not be throttled because a transport plen has been unable to be prepared. In parallel, planning capability should Indeed be strengthened. As explicitly provided for in the legislation, two or three Faritany capitals ehould already have the capacity (either locally or with outside help) to prepare transport plans on their own. Others would be prepared by MTMT on behalf of the local governments concerned. 9.17 The CTTUS, with functions and membership analogous to the National Technical Committee for Urban Development (CTNDU) described in Chapter V, provides another way to define the new partnership that should exist between Central Government, local governments, transport companies, and users. The CTTUS would be in charge of defining transportation zones, giving advice on the transport plan, and screening the requests for traffic rights submitted by private companies and individuals. The granting of traffic rights would be made by the President of the Executive Committee of the Fivondronana, after consultation with the CTTUS. In case several Fivondronana are affected, the Presidents of Executive Committees would need to reach a decision under the guidance of the Faritany President. This new structure should be experimented with as widely as possible, to get an idea of how local socio-economic, political, and administrative conditions influence the process and, therefore, how flexible its application should be. 9.18 Tariffs. There is no way around the application of fare increases to levels which cover the recurrent expenses of operators, plus an allowance for depreciation to sustain operations in the longer term. Discretion can be exercised, however, over the speed at which this objective ti reached and the measures taken to mitigate the social Impacts. Attainment of this finar.1ial objective in five to seven years for any surviving bus companies (cooperatives and independents would undoubtedly rech this point much earlier) would appear feasible. The social effects - 109 - would be mitigated if (a) quantity discounts for frequent users and other such provisions were offered; and (b) employers made better efforts to provide transport for their employees, which would bring a reduction in Government subsidies. It should also be possible to link fare increases more closely to the periodic improvements in vehicle stock, such as occurs when buses are acquired on a grant basis. 9.19 Road Imi,rovementa. As mentioned earlier, the highest-priority Improvements to the road and street network of Antananarivo have been identified. Since many of these 'IFL" investments (see Chapter IV) would benefit the urban and regional economy, a good case can be made for them in PER reviews. The case would be significantly stronger if Antananarivo improved its local finances and could bear a greater share of the cost, and also if a single metropolitan-wide entity were responsible for both urban and suburban roads in the Antananarivo region. 9.20 Metropolitan-wide Focus. As discussed in Chapter V, a single governing body for the Antananarivo metropolitain region should be put in place in the neat future (paras. 5.13-5.19). In order to determine the appropriate legal form this entity should take, much more experience should be gained with planning and carrying out sectoral programs on a region-wide basis. For urban transport this is in any event essential. The Traffic and Transport Bureau (Bureau de circulation et de transport, BCT) established under the PDUM project should continue and expand its operational studies of transport ir. the Antananarivo metropolitan area. Moreover, if transport companies operated on a regional basis, the more profitable lines could cross-subsidize the others and mitigate the fare increases borne by the poor. 9.21 Private Initiative. The 1988-89 laws have expanded the scope for cooperatives and especially independent operators to function as they have been for some time, but within the law. In addition to the legitimacy afforded to Independents, however, the legislation becomes quite specific on practices to be condoned or proscribed in particular situations. Applied too literally, these measures could perpetuate the notion of excessive Government regulation of the sector. In any potential conflict between the letter and the spirit of the legislation, it is important to preserve the spirit which is one of recognition of the value of private initiative and tolerance of a range of behavior by those providing transport services. Once again, the recommendation is that the measures contained in these texts be given wide practical application as rapidly as possible. - 110 - Annex 1.1 Page 1 of 5 AN ACCOUNTING FRAMWRK FOR THE REAL ECONOMY (SAM)t MAIR FEATURES 1. Analysis of the real economy has as its main purpose to identify and quantify the full range of transactions which take place within and between urban and rural areas of the country. As a consequence, the accounting framework used for this analysis must include relatively detailed balances of supply and demand by various agents of the economy, whose location is specified. The most practical tool for such a purpose is often the social accounting matrix (SMI). The SAN is a square table or matrix, in Which the rows represent Incomings (receipts) and the columns represent outgoings (expenditures) of all economic agents and accounts. Since the objective is to provide an aggregative picture of economic activity, the matrix should include all real and financial flows, corresponding to non- monetized as well as monetized production, consumption, transfers, saving, and investment. 2. The main structure of the SAMs presented in this report is shown in the Table below. Although it already contains 9-10 rows and columns, the different categories can be broken down further according to need. For example, In Table 1.4 of Chapter I, the 'households' category is disaggregated into urban and rural households, and Rgoverrment is broken down into central and local government. The 'wants' account corresponds to the final consumption of various goods and services by households and other agents. The other accounts identified relate to commodities (the balance between supply and demand), activities (the balance between production and distribution of value added), capital accounts (the balance between saving and investment), and the accounts of other regions or countries. 3. The property of SAMH to juxtapose economic activity in a specific area with that outside the area implies that they can be constructed for a city or region as well as a country. A national SAM would distinguish urban areas and rural areas of the country. A localized SAN would show transactions between one particular city or town and its rural hinterland. 4. In the first case, that for all urban areas, the SAN would show (i) the exchanges within urban areas, forming the urban economy; (ii) exchanges between this urban economy and the rest of the country (i.e., the rural areas); and (iII) exchanges between the urban economy and the rest of the world. Similarly, a rural SAM would show the intra-rural transactions, the rural-urban transactions, and exchanges with the rest of the world. These two SAMs identify the various components of urban-rural exchanges at the national level. S. In the second case, that of a region of the country centered on a city or town, the SAM Identifies the various transactions of the city or town considered in relation to its hinterland. The SAM compiled for the Antananarivo area is described in Chapter 1. Annex 1.1 I I ~~~~~~~~~~Page 2of 5 fil ' . ,., 1wI i,. -'---i-[e- sli1 IXk0g Ii:!§8 g 3 i t I I 0 5 . , g g I I I I g t 1W~~~~~~~~~~~~~c { ;§§{ t,gg stl. . . - -g -- - .. -l - - - .ALis, .iI lIL 11W! !|!.. I~~~~~~ l *--tsp iou#t't n01 us ea usomw lO 001 £00 n SsIU* WULt" ea wt ma _ t CS miS mS at ma mi aa. £,4. so "a ooalupsawals oO9goingPpow mus"Of. 65 011 o"IstOUaftn4u 001 00 1tilPo *"$Om Gin t 11 mUtton wto--- -t s.ta 'TtP v_I4u _ _I _ _0 _ _ .. .-. . ..p .w~ _ _ . _ _i _ _s~~ _ mini m~~~~~~~~~P fM_o u" jMm m.LoaUU (an s I (eM (,e0 (ti/si ) N W) ew/uf) mu mu mum (moe osg.aa um u u o _9 otm is O 05U0.S 5on 600U 00w30dlWjSEH *iow Ofe-i050U 60-i LL Um gm'0 MWiSlSIW l8 ususui an "a "*-ifa S goo"N uwiiu inn am Vnts go-I *st- Wse/uI mum ((ha/so) m smasutS swim IOU~ 63avm (a~SmLUSp EVUft UIIO O SDVEIEu 3mm t o ls" EX-COSMUNES URBAINES (25/46) ANALYSE DES RECETTES TOTAL FIVONDRONANA FIRAISANA EVOLUTION DU COMPTE ADMINISTRATIF DE EX-COMMUNES RX-COMMUNES EX-COM MNES 1987 A 1988 URBAINES URBAINES URBAINES (25146) (6/9) (19137) RECETTES PROPRES Impots directs 702 801 122 Impots indirects 51 22 20S Revenus du domaine et des services 191 201 14S Produits divers et accidentels 242 272 112 TRANSFERTS Contributions, ristournes... 371 331 502 Recettes additionnelles extraordinaires -442 -481 -82 RECETTES PROFRES 41S 471 141 TRANSFERTS 241 191 45S RECETTES TOTALES 301 291 341 (HORS PRODUITS ANTERIEURS) PRODUITS ANTERIEURS Recettes additionnelles ordinaires 871 881 862 ANALYSE DES DEPENSES TOT?t, FIVONDRONANA MRAISANA EVOLUTION DU COMPTE ADMINISTRATIF DE EX-0MMUNES EX-COMMUNES EX-COMMUES 1987 A 1988 URBAINES URBAINES URBAINES DEPENSES DE FONCTIONNMENT 371 381 332 DEPENSES DI'INVESTISSEMENT 991 1011 951 DEPENSES TOTALES 451 452 432 AUTRES DEPENSES -371 -421 43Z ina=55wamwa=5=wamwwinmauuiwa m.uuwmunmm flI.IItU#nl311 S_ SIWIIIt0 I - 129 - Anrnex 6.3 I Page 3~ of 3 D I * X l V 5W " l' I i 1S#I I xI a 'I , lWEs W!I!I ' I lliII!II 8I 5 ' | II| i i ' fi, S! I !|f18*| 1 i~~~~~~~~ I j 1'1%l 8 lii II t8| I# II j |I ' I I, II i i0||i EI - 130 - Annex 6.4 Page 1 of 13 (ANALYSIS OF BUDGETS OF PARITANY) ANl&YSE DES COCPTES ADMINISTRATIFS DES FARITANY - 131 - Annex 6.4 Page 2 of 13 (ANALYSIS OF BUDGETS OF FARlTANY) (INALYSIS OF RECEIPTS) ANALYSE DES COMES ADNINISTRATIFS DES FARITANY ANALYSE DES RECESTES FARITAWJ9 COPTE ADMINISTRATIF DE LEXElCICE 1987 TOWl Ante.- Antel re- Fltm- Mbejang Toegstna Toltaro (MIll Io Io FdFM) nmrl,o n"as Caet" -- -- - - -- - - -- --- - - - - RECETTES PROPRES FONCTIHEMENT Roven.. do blowb unit.. do production a 4 10 20 0 a Produlth timcl. 1 1 Prelevometo tat txes specltiqus 217 98 49 51 23 2P topots at taxs lovee our l-* terrteir 2438 1611 10 194 229 294 U1 Produits .xceptliowloo 65 6 1S 34 1 2 0 INESTISSEMENT Roecuvrement d. prte/ *Vece 32 28 Al ention 1 1 TRANSFERTS Ristormeo at rever _ema low44 4726 401 2811 739 1529 1171 Recouvreate et .s.bveentloe C6 12 is 28 0 PROJITS ATMIERWS FONCTIONEMIT 2822 128 8 276 am 8 O 42 189 IWESTISSEMNT 38 14 6S8 1442 1982 013 159 RECETTES PROPES 3g6 15 16e 22m 2980 1146 183 TRANSFERTS 110 4787 401 23 739 1555 11?? RECETTES TOTALES 1465 9883a Su 2816 1019 27?1 1310 (HORS PRODUITS ANTERJS/ RESEVES) PRODUITS ATER RU S U60S 1280 84 266 100 665 299 RECETTES TOTALES 21216 7s13 1481 5117 230 3386 16"6 fRITMW 02/OS/0 COMPTE ADMINISTRATIF E LEXERCIC M 1938 Tot4l Ant.a- Antsire- Flone- rhbjens. Toamasina Totlere (Milons de Fll) neilvo "no nantee IIECETTES PROPRES FCNCTIOUNEMET Revenus dem bl t unites e prodction 19 0 t 3 2 2 Produits fineaclws 2 2 1 Pr,loausents t taxz* pectlfiq. 32 49 68 S2 21 27 lpots et taxes leye our Is tarritoir. 2427 1166 l61 284 276 480 149 Produits oxc.ptioonnel 25 10 2 0 1 2 4 INVESTISSEIMOt Recouvremnt de protO/ evece 9 Al oiention 0 TRAISFEETS Ristournes et rsevromnt 1in" 4910 Us6 32s 7t6 2014 1666 Recouvrssnt, et subvatiols 228 184 19 0 22 0 PODUITS AERMS FNCTIIONIEMET 8283 1767 203 1il 246 14 a0 INVESTISSEMENT a446 407 15a8 417 S86 19 ____________________ --__ -_--- ---- --- -- --_ -_--- -_ ,. __. ______ -____ ---._______------ ----- IECETTES PROPRES 207S 1239 229 28? 327 423 178 lRANSFERTS 13281 g1as 585 am22 756 2M37 1601 RECETTES TOTALES 16967 08 764 n6o 0113 2466 1736 (RS PIIOUITS AMTEURWS/ ESERYES) PRhOUITS ATERIEIES 0788 1757 t07 2642 01 694 210 m RECETIES TOTALES 2290f 33? 1434 0122 1746 3359 1944 0 - 134 - Anexc 6.4 Page 5 of 13 X I #| | §I U!f1 Hg I § "l St 1'T 1' i~~~~ ri ' gl rlZ 4.j1}18 i 1 igl P8111M 12/13/13 TO"i AI- Ast Fee j lm1e ToOter uerlw. es ret. IW*STITIWOn VCIT FM GUNK(11 nrTE s - s eves. bles s mel* de prdmotlo.. as 6.23 .3 0.1 0.13 9.16 Predelvt. fleesmel a3 6.13 9.13 6.13 6.13 C.1 6.1 Pu.leveeeee Utse opesfqme is 1 9.1131 6.13 6.1 *." 2.23 IRete .5 team l~ Mr le tweltoter 173 28.1 1." 7.4 M2.43 U.1 1.3 Pelit *ea l.e1s 1 9.1 .71 1.1 .13 9.13 6. IMESTISSEUDT R.eAv reamb de Prow/ ax*e 1 *.1 9.3 *. 0. 113 9.3 . Aleestoes a3 6.413 9.13 6.13 6.13 610. 3let.u,see .5 r.memsmte 753 7. 1. 13.11 72.3 5. 13.43 R.amvemet..5 awbesatlO.. 1_ M .23 9.1 0.43 0.13 1.13 6.1 P3651111 AMSIBS FOKTIONSEI_ 1IX ARM 47.153 4*.43 8.1 S.A WA.S fOltI_tEU 19. .0. 6.13 3. 22l.7 U."3 mICEvES rac'3eS 2to 26.2 8.113 0.73 21.6 42.43 10.11 TUM TS 713 14.10 0.S3 726 67.41 13.1 (mm misMMMm N-n- IlECETES TOALE 11 .13 W." 11.13 13.13 1.13 80.1 ( fU35 P61311115 M75WD5/ KUUVS) W1 TS AmuIEw 46.TX 1.23 144.53 13.41 19.1 24.831 22.M o0. iPltIT1Nt 12f20/09 Istl Aatsua Aetele Ftse UbJ_eo t.soe Teflere &EPRTITION OES RECETVES PAR OI4O (1980) tEC11tES F6PrS VUNCTIONEWIT Rewu,0s dbe bm at amUS do prdectioU S. 0.9 1.5 0.11 t . .1 .U1 Prodult. tlueseIl.e 0.95 L. *.2n 0.9 0.95 0.. 1.11 Pplg_emgm" S seame .s _11qm 1.1 0.91 0.95 0.95 4.CO1 O.I 1.05 ImptOe S tax" tons SW 1 mrIt.Iro, 16.21 16.45 21.11 7.01 25.41 10.55 0.05 Prodult e maepto.m.ml 0.21 *.Ss 0.511 0.95 0.111 0.41 0.211 IWvESTISSEMT Recvr_mt do pro%*/ gev095_ 8.11.9 0.03 eU 0.*9 *..9 0.01 Al I.Stioto 0.95 6.01 0.95 0.95 9.95 1.11 0.95 mestonwes .t revesumesaft 010 71 00 1.71 00.21 01.15 95.75 Raco_vrsme o. sub-stleas 1.41 2.65 0.95 0.55 0.05 0.61 0.41 POWUITS ANT0IS FONCTIUEN T 20.95 27." ".211 ".2121 21.05 0.5 8.45 IStUSSET 21.9 0.. 01.11l 42.61 3.5 6.n l." -_ -________ _- _---- - - ---_ --__-___--___z ___ aCESrn p,rj'g1j 1ff0. 19.41 80.03 0.95 6.21 17.41 10.03 tIANSFERTS .2 .1 70.6 "2.11 Wa0 02.1 95. aCTSTU touALES 195.03 1. 1.9 1 111.11 91115.95 1,1.41 1M5.0 (N5 P3001117S ANTEIEWSS/ SESEAVES) PlIwars Ajstmt 42.21 27.61 07.05 71.11 01.11 86.15 12.11 8 K o ¢ - 137 - Annex 6.4 Page 8 of 13 AALYSE DES COtPTES ADMINISTPATILS DES FARITANY ANALYSE DES DEPENSES FOITMf cEU MIINI TI DE L'SMCE 167 t.l Lt. L Atelr- Flom- IMIJ toml. toll ( ell)t de Fl) "erive MOSS rots" Frei* de perasoeol gm2 5277 578 1o1 72 u2 1148 h.ra *t tot.. tet 1o.usobles 425 156 4 6 6 6 U Treveux at wot.. eawmro 171 0 25 24 2 2go Frle do stle.gaael* sM 1 221 24 17 75 47 24 CoatributIe. et eubve.tle se1 546 26 45 17 46 so Caerge .exapolenlles a Astre ster (1) 46 4 I G ESTIISS EMIEMllT Freft 9rtaordlealr 5 146 166 2 2 Pn**/ *Vn_* o@ at Uter on P* tletelpa. a de tr . d.'qpett_ tll 61 1146 22 71 Im l li t)o.s 2414 512 245 12 647 s I EXCOT CAPTALZIE 157 so 811 54 2U 6 169 D£MSE.S mm F(ICt1 4 2 2188 65 164 12 OEP3S 0' 4#9 1694 483 1101 871 711 169 DEPEISES TOTALES 16475 gm oil 8294 1426 265 1455 (HOSS EXCEDENr CAPITLSE EXCEE CAPITASF.S 157 IN 811 584 253 IS , O. iS mOTALES 1084 i5 124 825 1084 1614 0 (1): Fee ; tf looealer* lqst. t texas, domaas"-Igt4Nv, aebss go¢t amrse Sterlegre W FMITAM ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~1P/ C 5WM AINITI E LK E VORCE 1 T. Ante- Aei- Fl-t_ J T lr (IIIIIlow deFtl) erttivo se r_ d" WmtllCtI_t SIIECT Frele de prsosel 1is 32 446 231 725 26 1417 Ders at fosraltt c bl 67 152 76 is6 1 of so TroVeu et m crvles .erlours 201 156 16 6 26 27 U Frei d. Ngtion _r abl 814 9C to 22 as 6 3 Coltributio. dt ub.tl 7# 024 2 1 2t 66 47 Carg. .ucepto aell Om 54 Aute eberg (1) 5 1 8 Fceb setrerleeue 1U 4F 1J3 £ 2 2 Feelcz* 10t" *di *im 2 P.ote/ eve.... a leas St msr.. too" 0 Partipation a dw trevex. dm "1s1 t 20 407 12? Ibi l..tl... 1524 452 106 S 3 2U 140 EXCEOT CAITALISE 217 2 #31 1251 256 166 1S WPSES OE F"CTIEMB 12 4765 Su t6o 97 220 154U OVFENSES D INVESTISSEOENT 986 436 1 t 626 1tS DOMM~~~~~~~~_ WALD s _ ion _ w 2mIo DSeESs tOTALSs 19421 5772 190 869 1335 2822 198 (HORS EXCET CAPITALS) EXCEDENT CAPITALtSE 2176 2 831 1261 21i 1i6 16 w m m m i n -~~~~~~wsom Nminw ~ = wmuus - immumsoamm EPESES TOTALSS 1"67 5774 1i6m G 16is 27 16 o 0F (1): Freis fleoucebrs lupt t ter _re -lnt..e .bug ar xerctus eote.bs_ - 140 - Annex 6.4 Page 11 of 13 I U!! 3ll Y U i$ IUl . l I I I £ I§ I I!'! I I!I I_ ill I I I I ; ll ill ! 1 l! l i FARITAN ( reSttes toteil) 12/IS/So Total AMt..- Aeit*lr- Flene- Usb.jeRM Tosesele Tolierr SERMTITIUN OID RECUTIS PM 131IUg (1W?) ~~~~~~~~~~ - - W861 FONCTINEMET DSICT Frei* de persomel 511 on 4U NS 201 01 711 .orws to.r.lts_ seesmsbloe 81 8 a 5 23 41 Ss 23 Trwavtx et wvkss ex%ewiuse Ax 1i 8s Ax 23 1x 11 Frets d. estiom .El 23 23 83 1i Ss 21 21 Coatrib.ti.met s.bmttie 4 i S1 iS is 2s1 A Chtaro *xceptioesl fe is1 is S es S a a Aetro charge (1) a a o 6s o n INVESTISSEMEN Fr*i. . tr.ordleairm 23 81 2W 0s a d u Prots/ *Vmcm a loas at us,- teni 4S 1W es a U U Participation a doe t_wese .lpinst u U U i 2n ex I l_biIltet 1o Ss go of11 251 113 IEXCESET CPITALUE so IX 8x ll 2S uS DSPEGES DE FINCTIUUEUT eJ8 723 0IS 0WENSES 0'DIISTISSEE1 2 27l3 403 011 2ex ----~~~~~~~~~~~~ew U--- m -|s_W BUSSES MM" _ _ U - Urlt (HONS EXCES CAPITALIS) EXCEOEST CAPITAUSE Us l31 lox 21 111 (1): Freis fI s.nl. ptso o 4 , d s-1, Ar.s Ic. in est - 142 - Aue 6.4 Pa8e 13 of 13 I II UIul ll | '" I I "" 'I'!f Ull j } Ut| I I31ll ll# £1 1 I 3E!UEUU ""In 'l0ni VI ,~~~uuuuu !~~~~~~~~~~~U U jEUII 1 I~~~~~~~~~~~~~ - 143 - Anmex 6.5 Page 1 of 7 (ANALYSIS OF BUDGETS OF FiONRORANA) (FORMER URBAN COO!NES) ANALYSE DES COtES MINISTR&TILS DES FIVODOtN& EX-C!NES URAINIES LISTE DES FIVONDRONANA ux-ccius uu&ass ANALYSES (*) VARITAN! FIVOIDRONAN& FIVONDROINA El-CU ANALYSES (9) (6) ANTANANARIVO ANTANNRIVO-RENIV. I ANTSIRAIE I X ANTSIRANANA ANTSIRAI ANA I NOSY-RE X FIANN FIANJLSOA NARAJANCA KABAJNGA I S TOQASINA TOANASIU I S SAINY-KARIE x TOLIAR TOLIAR I (*)s Leur population represento 1 170 000 habitants (estimation 1988) solt environ 80X de l'9nsemble de la population des Fivondronna ze-Comunes Urbaluae i1 FIVORClNAWPOKONTAWV WRBAINS (0/9) 02/06J91 ANLYSE DES RECETTES CUWTE ADMINISJAMTIF DE L>EXE11CICE 1987 Total Antan.- Nosy-S labhJ ong S*int.- TYommlun (Million do fP1) Renivoh. Antstirb. Mrie, RECETTES PROPtES Iup.to directs 904 469 4? 28 1I1 1 21 impoe indirct. 283 128 28 13 48 0 71 Roevue do domino et daso merI. 028 333 a8 40 38 1 129 Produits divrer .t acldeln 43 37 1 o 2 e 3 TRANSfERTS Contributions, ristournee... 24 Si8 263 i7 222 27 326 Recette edditionselles extrordlnsirm 6S4 258 01 161 19 1 34 RECETTES PROPRES 1i68 on6 158 e1 238 a 412 tRANSfERTS 3194 1is" 34s 234 241 28 359 RECETTES TOTAP ES 6sg 2965 602 N81 478 80 m o (PMNS PROWITS MTERIEUlS) PRODUITS ANTERIEtRS Recettes dditionnellea ordlalres 426 is 20 0 263 1 116 ANALYSE DES DEPENSES COIPTE ADMINISTRATIF OE L'EERCICE 1987 TO"tl AnsSe.- Nosy-Se MahaJanp Seinte- Toaasian Renivoh. AuttIrabe Marie, DEPENSES DE FGCTIOmEMENT 3326 2145 360 182 332 20 342 DEPENSES D'NIESTISSEMENT 409 140 18 20 224 4 4 DEP£NSES TOTALES 3783 2286 $74 iNt 6S6 23 346 W AUFPS DEPENSES 76 443 76 96 25 1 124 .~~~~~~~~~~~~~~~~~~~ _cce== _ 3s==s_==== - 146 - Annex 6.5 Page 4 of 7 q| q' _ na X ifW : i: 0§ f | 4I I If I"1it I I alw.- idX §|p; xXl§ii 5 ra a 01/111t ISL KES IECfTE Totl EVOLUTIO U COMPTE Mh TIF DE A.1.- Sbeq-S Me g RSetl- Toemmele 116? A low R b. Aietalbe "Sri Iqot. directa so 112S li1 261 a 1821 61s Iqet. ledirocte is as -141 as -61 1121 41 Rove. do da_m. on 22 2m -151 111 21m1 51S Proadita divtr .1 scdesttle 211 -161 121 411 1861 88281 86 ItASISFERTS Contributiose, ritose... IS 221 6so six I6 20S 28 Recotte altikme m el e otreo"le.re. -461 -41 -761 -211 -8S6 ON 461 RECETlES RnWR 411 US 14 2S -21 41381 61 TRASSIS to 1n 2 -6S 071 261 1 AECETtES TTAlES to 27 241 -6 461 US 42S (NM3$ FUITS hIBaUS) becott. adiut_ol.s. ordlmeire. s 621x 145 S1 -44S 485 2811 ANLYSE ES OPIESN EOLUITIIN U COWtE AU1STRATV K To"$l Ant.- Is_-". ubs st.- T...i" 1917 A 166 IRel ob. t eMario ~~ in m 0in --- --- oummm OEPENSES K FO5CKTJ11EN as 8 281 6 14S 781 7,1 DEFESES 'DI ISSEt I 16Ms am6 -71 441 61S 61 3x i Dusan MALES 461, an u2s us 5 2 to AUTES WO S iii -4m uss - fs ex FIVO _DROPIAWW.TAN URAINS 81/11/98 (= dm recattes toa") AIALYSE DES ItECETTES REPARTITION DES RECETTES PAR IGIlE (19V) Total Antan.- Nony-S. fabajenop Saint.- Toama si Realwoh. Ants) rab Mario IECETTES FRES 1ipta directs 181 16X 9o 9X 32% $1 271 Impote indirecte 61 43 CX 411 lox Is 6 Revenue do domains, t din services 123 113s 181 l3x 8a 3 X 173 Prodmits divers et accideatels IX IX oX as ox ox oX TRANSFERNS Contributions, ristournes... 821 593 621 28x 461 86X 42X Rmcettes OddKonsolles eatraerdinlrem 11X 9s 161 61X 4X 4X 4X RECETTES PROPRES 373 33X 321 2X 61 8 61 531 TRANSFERTS 6SO 7X e88 741 6t 928 4711 _ _ 5__ ==8=~~~~~~~~~ ~ ~~~=== _= 5 _== ========: PECETtES lOTALES 1Am3 1811 103 18 I" lolw 1o7 (NORS PRODUITS ANTERDEtIS) PRODUITS AMNERIEURS Recentes addit;onn le ordina res 8t 11 4X 21 5X 4X 1SX ANaLYSE DES DEPENSES REPARTItION DES DEPENSES PAR NATURE (1987) Total Antan.- Novy-Be Watajanga Sainto- Toaasaina Renivoh. Antairabe Mari ==__==~~~~~~~~~~~~~~~~~~~~~~~===G =-= _G== =1:==# ===z==== ===-=S_=== DEPENSES DE fONCTIOMEMENT 8BX 94X 95X 871 s8x 84X 9SX DEPENSES D 'I"ESTISSEMENT 113 as 6X 13X 4BX lx Is DOFENSES TOTALES 183 183 18i3 1S less lax1 18to o c AUTRES DEPENSES 29X 1i3 21S 68x 4 S 32 '6 - 149 - Anneg 6.5 Page 7 of 7 *- I I-Il|1l1| |1 8~~!U RUJI-"-|1W! . t1r:; II SX f|uu| a 1| =3~~~ ~~ I '#1 - |11111 '}~~U! f0:u if' , |l , ta -150- Amex 6.6 (ANhLYSIS OF BUDGET OF FI3&SAN) (NORE URBAN COMUNES) ANYSE DES COgmTE ADInNISIFS DoS F EX-cCImS UMAUS - 151 - Aunex 6.6 LISTS DES FIMISAA EX-CItIES Page 2 ,f 7 UMAINES ANALYSES FARITANY FIRAISANA 13RAISANA EX4CU ANALYSES (37) (19137) ANTAIANARZVO AMBATOLAMPY AMBoHnxmN X ARIVONIMAKO x MARINIARIVO X SSIROANAWANDIDY ANTSIRAMW AMBANJA AZBILOBE ANDA ANTALAHA SAMDAVA FIAMARANTSOA AtBALAVAO O N ~~~~~x AMBOSTR IROSY I _SAX A x NAZIAL RA t MANANJARY VANCAINDRANO X MAHAJANGA ANTSOHIHY X SEFANDRIANA NOlD MUEVATANANA X NAINTIRANO NANPIKON? MANDRITSARA MAOVOAY X PORT-8ERGE x TOAMASINA ANIATONDRAZAKM X FENERI VE-ST MAROANTSETRA KORAZANCA X TOLIAY BETROKA WAMO MOROMBE I X MORONDAVA X TOLAGNARO _____.___.____.___ ------------------- ----------- FDAISAM'NPGKOTJAWY IMUIS (19/3?) 2/06/W ANALYSE DES RECETTES Total Antenn- Ant Itor- Floea- nobajangm Tmo.nlm Toel Ire COSWTE ADBINISIRATIF DE L3EXERCICE 1907 naruio Dane reoteos (MIllions de FUD) (8/6) (6/6) (6/J) (0/8) (2/5) (2/5) RECETTES PROPRES impot. directs 158 63 n.d 20 41 29 7 lapots indirects 6s 6 n.d 2a is 0 18 RUov ue do dombino *t ds ,rmvices 204 29 n.d 46 47 09 12 Prodilts dIvers *t acIdmiamtI 18 1 a d a c J TRANSFERTS Contributions. ristourns... 748 167 n.d 267 106 101 97 Recottes dditionsells sxtraordlinoir 6 6 C.d 1 10 88 6 RECETTES PROPRES 485 89 n.d 11 O64 100 82 TRANSFERTS 803 178 m.d 267 124 143 0? RECETTES TOTALES 128 262 m.d 86? 228 2C2 129 (OMIRS PRODUITS AlI2ER2I1RS) PflRDUITS ANTERIEtS Recettes *dditionn.llen ordineirss 219 18 n.d 111 4S 84 0 ANALYSE DES DEPENSES Total Ant"na- Anteira- flans- UaheIoaga Tormeiao Tollore COWTE ADMINISTRATIF DE L'EXERCICE 1N? narivl no n r atet (8/6) (5/6) (6/8) (l0e) (2/5) (215) DEPENSES DE FONCTIOWIET 80? 168 .d 2S2 1 DEPENSES D'IESTISSEMENT 105 84 a.d 48 40 40 9 DEPENSES TOTALES 972 202 n.d 80 196 176 90 AUTRES DEPENSES 49 10 n.d 18 8 11 6 FWAISAIPWKSTANI URBAINS 02/05/90 AALYSE DES RECETTES 'otal Antana- Anteisr- FPlan- Majenp Tomostis TWlinr CUWTf ADMINISIIATIF OE L'EXERCICE ION narivo nn* rente"o (million de FPU) (8J6) (6/6) (60) (0/8) (2/C) (2/6) NECETTES RWS Ipwt. directs 17S a .d 82 82 82 It lopoto i ndlrote 78 7 n.d 26 1S 18 12 Ibews do domaino t des ervices 282 at .d 6a C S @0 12 Prociuts divters et .ccdmntels 14 2 *.d 8 4 6 1 TRANSFERS Contributlons, ristesross... 111 221 n.d 88 218 161 104 Recstt.s addUt.oell Istreordinsirs 680 a n.d 1 1 U6 0 RECETES fRWRES 4# 869 m.d 114 O1 188 87 T SFERTS 1168 224 .d 88 21 217 104 RECETTES TOTALES 1864 888 n.d 5607 U8? 80 141 OaS PRUIrS ANTERI) P6WI1S ANTERItJS Rnettss addtd,mwIIes ordinaires 488 n m.d Ift t9 0 88 AMLYSE DES OEfESES Totl Anten- Antairs- Fil_- lahajeng Toameenl Tollers aW1E AWINISlRRATIf DK LEXERCICE 1988 nerivo mes, renteos (8/6) (5/6) (6/8) (8/8) (2/6) (2/6) ~~~~~~~ ~~~~~~~~~~~~~ ~~susmmm. -=Oam1M=N DEPSES DE FCNCTIONEMNT i7l 2t2 n.d 811 267 191 110 DEPESES D'KINESTISSEMEIT 828 *8 s.d 116 56 71 32 =5= mn= =77 7T 7 K ATEPES DOENES 1894 8 n.d 42 267 262 142 AETS OEPEWSES 70 is n.d is 14 is 9 FUAISAWU(SSAIS Umalls 40/1111,u r _ ~~~---- -----9 .LTSE DES IECETKS ToWSl Aot...- Lotem- Fles- J Tossesle Toellotr IWLMTRU OU COMMI WOUDIZ1TIF KC morive no"e roots" 1? A 19A 3 (/5) (5/8) (186) (1) (2/) (2/5) RECTES fURES loots difects 1211 an .d 141 -22X 1 Impote ladireat 2an 6 .d 1to s1 Roves.. de d_a1i.. ot d m .vims 141 1ll a.4 141 171 d s Pro.ddts divors ot .midmmt.1 11S 9n s.d a -2 so 871 ThANSFERI Contributions, rAtmavmas... 5 1 s.d 411 1051 is 71 Sacettas adti.as.l leg outrdle tee -l -451 s.d -171 A 4 RECETTES FWE 1411 WI .d 15i 21 to 141 SF 453 s n.d 471 11 611 s i _ ""e--|--- ~ ~- - - ~ua RECETTES TOTALES 545 27 MA.d SI 461 m73 ON (HOmS nMWITS A_TI935) beOMte adilti.I le ordlastree s6o 53 m.d 425 75111 161 ANIALY5 ES EPESES Totol Mtwre- Aetere- Flets- MIbht)., T.a.la Togte or EVOfl1t OU COIrlE IIIIII EATIF OE artwo VAns reet... 1967 A 1S81 (a/5 (/6p) (8/5) (C/8) (2/5) (2/5) ----~~~~~~~~~~~~~~~~~~~omm m 3nom uomumm ininamag DOUES DE FlGlNMTI US 0 s.d 1a1 ax 46 so6 DAEUES D'P ESEISE461 5ox s.d 1745 243 7931 5i WEIumISES 451 -.' mJt$ .d 415 51 491 611o LURS DEP>EISES 51 66 m.d 141 65 451 571 Annex 6.6 -155- Page 6 of 7 S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~S t - .4 #4! 1|1§ II R01 * a.i-J .4 #4||SS! a I a a d il I rI |1 SW 1! X [ ilS1 1 |r S~~~~~~~~ -0 l ,s II IiI'I| -I56 - nex 6.6 Page 7 of 7 l AU! S|II I *1 | i :1 ! !3iuI l1 I 1S1 l U~ || U|#l U 1l K 1-!"|1 !I0:|1 *3 1- t11 I 1| |~~~~~II I1:Xi,; WEUZtE1|l - 157 - Annex 6.7 Greater Antananarivo Size Distribution of Land Plots ---------Distribution (2)---------- Fivondronana Number of Plots Less than From 1000 m2 More o1000m to S ba than Sa Renivohitra 96,000 90.0 10.0 0 Avsradrano 172,000 76.0 23.5 0.5 Atsimondrano 139,000 68.0 31.0 0.2 Ambohidratrimo 74,000 76.0 23.0 1.0 Arivoniuamo 4,000 25.0 65.0 10.0 Total or Average 485,000 67.0 31.0 2.0 Source: AURAIGroupe HuittBCEOM, Etude de Developpement du Grand Antananarivo, 1985 Annex 6.8 -158- Page 1 of 2 REPOBUKA DEMOKRATIKA MALAGASY Taniudrazana-Tolom-F; avotana-Fa- af ahana mtNINSTERE DES TRAVAUX PUBLICS MINISTERE AUPRES DE LA PRESIDENCE CHARGE DES FINANCES IT DE L'ECONOMIE Burnu des Projeta Urbclo Dectios deo Tri6 I) E C R ET n- 87-287 portant cr6ation du Fonds de l'Infrastructurc pour I'Habitat Urbain (F.l.H.U.) LE PRESIDENT DE LA REPUBLIQUE DEMOCRATIQUE DE MADAGASCAR; VVu la Constiton; Vu le decre a 8-W 9 do 13 d6cembre 1984 pou rsaicadon de rAccocd de Cr6dk au Developpement n 1497 MAO coacu le 3 juillet 1984 entre la Repubique DImo"isue de Madagascar et l'Association teronal pour le D6veloppeme>.. relaif au Projet de DMvelopenment Urbain de Madagascar (PDIf); Vu le Decret n 84-071 du 15 fevrier 1984 poa creation du Bureau du Projet urbSin (BPU). Sau pooposition conjointe du Mlnisre des Travax Publics. t atu Ministre aupres de la Pr6sidence charge des Finances et de l'Economie. En Conseil des MiDismres, DECRETE: "cle pmier 11 est crt* un Compte Special d*nomme wFouds de l'Infastructure pour l'Habitat Urbain" dans les livres de la Banque Centrale et au nom du Tresor Public. Aicle 2: Le compte sus-vise est alimente en recette paw les rembounements des ptts accordis par leTresor Public aux Collectivites Decentralisees et a des orgaismes chages de rex6cution de progammme do d6veloppement urbain. Ce compte pounra recevoir d'aute concours financiers incerieurs ou exteriours relatifs i des programmes de developpement urbain approuves au plan national. Ce compte compottera ainasi des sous-comptes correspondantsd ces differentes sources de financement. jicl 3: Les versements s'effectueront conformement aux echeanciers des pres concernis. Le Tresor Public informera penodiquement le Miniscere des Travaux Publics de la situation des comptes et sous-comptes du Fonds de l'Infrastructure pour l Habitat Urbain. Aricle I: Les modalites do fonctionnuemet et d'utilisation du Fonds seront fix6es pa arctte conjoint dui Ministre aupres de la Presidence chage des Finances et de l'Economie et du Minixue des Travaux Publics. Annex 6.8 -159 - Paego2 f2 =,:de S Des an*ts d'applicaiosi ptciserou en nt que de besa les moddites teciques de vesoneni et d'alirentation des sous-cmpS. Artide 6: Le Maistre =pres de laPr ldeace cbat des Fnes et de l'B_O¢oml. Ie Minre des Tvsux Publics. le ?Miisw de V'lnt6ieur. Ie Mintie de rindusuie. de alEnegie et des Mines sewa cbegd chum ce qui les concem de lrecuion di peumi dfe s qore pubmih au Journ Officit doe l Rpubfiquo Dftoaique de Madqascar. - 16C - Annex 8.1 Page 1 of 5 THE URBAN HOUSING MARKET A. Housina Situation and Recent Changes in Housing Conditions 1. d&-sing conditions In Antananarivo and in the 6 other major cities of Maw¶agascar depend critically on the rate of population growth. Antananarivo, with 690,000 inhabitants in the center city (Fivondronana of Antananarivo-Renivohitra) and an estimated 1,025,000 in the metropolitan area, is growing less rapidly than secondary towns. The housing stock can be estimated at 171,000 units, of which 452 are located in central Antananarivo, 262 in the inner suburbs, and 13X in several small centers separated by rice fields within a radius of 10 km from the city center. However, Antananarivo also contains two-thirds of the urban population, and in the not very distant past has grown faster than presently. The result is severe overcrowding and inadequate services. Houses are generally very small, with one- or two-room dwellings comprising 801 of the total. These dwellings have an average of 5 persons per room. Fully 602 of dwellings have less than 17 m2 of livable area. The small size of dwellings is more pronounced among low-income groups, but affects the great majority of households only 202 of whom have more than 35 m2. Although only one-fifth of the housing stock is considered to be very poorly constructed, housing quality is deteriorating over time for lack of adequate maintenance and with the increase in density. Poor access to public service networks is commons 3 out of 4 dwellings have no water (8 of 10 outside the city center), and half have neither water nor electricity. The situation of renters is marginolly better ,then, tbat. ofownerss. 262 and 20S, respectively, ace connected to the water network. 2. The 5 other Faritany capitals and AntsirabQ, on the other hand, have seen a more rapid increase in population since the start of the 1980s. Without enough serviced lots, this growth has led to the appearance of squatter settlements on public lands. In Mahajanga 402 of the population lives in unrecognized settlements. Although to a more limited extent, Antsirab; also has squatter areas close to the town center. Local authorities in Antsirsbe are working to organize these areas from their limited financial resources. There is also a demand for rental properties from employees of local firms, but the market is tight and rents are relatively high. 3. Renting is in fact the most common form of occupation and the most affordable to practically all groups of the urban population. In Antananarivo 602 of people in both the central city and the inner buburbs rent their dwellings, as do tvo-thirds of medium- and high-income households. A majority of non-renters are found only among the lowest income groups, living illegally in rudimentary shelter. A significant proportion of Antananarivo households (152) claim to have "free accommodation," but the fear of paying property taxes may influence these declarations. In 1989 the average rent was FMG 11,500 (US$7.50) per month, varying according to location; rents are 30-402 higher in Antananarivo- - 161 - Annex 8.1 Page 2 of 5 Renivohitra than in the inner suburbs, and 20X higher in those areas than in the most distant neighborhoods. Average rents have increased 50S in nominal terms since 1985, less than the rate of inflation. 4. In practice, given the demand for rental space, owners have more opportunity to rent separate rooms and obtain an adequate return. However, they can increase rents only to a limited extent because of the prevalence of space offered for rent. The limited rent increases are offset by allowing the quality of the buildings to deteriorate. Typical monthly rents for dwellings with less than 12 an of livable area are FMG 10,000 in Antanarivo-Renivohitra and FMG 7,500 on the periphery. A good quality two- room house rents for RMG 60,000 per month or more. The lowest Income groups devote in excess of 15% of their income to housing. However, this percentage falls quickly beginning in the low-middle groups, mainly because the very poor housing conditions force them to live in dwellings worse than they can afford. At the other end, a speculative market is developing in rentals to company employees and to expatriate advisers that is displacing middle-income renters. 5. The supply of publicly-financed rental dwellings is only a small fraction of he total supply. Out of the 95,000 rental dwellings in Antananarivo, only 1,500 have been built and are managed directly by the Government. They serve a small minority of the 27,000 households of civil servants (excluding military and 'eachers' housing). These dwellings are typically more than 30 years old, poorly maintained, but because the rents are close to zero (FMG 1,500 or US$1 per month for 'low-cost' housing), there is little turnover. 6. SEIMAD manages 3,156 additional dwelling units in its four Antananarivo developments 1/ following the sale in 1988 of 1,173 dwellings. SEIMAD units are open to all groups of the population but given the low rents, turnover is again very limited. In conjunction with a policy of improving service quality, three rent increases, of 10 in 1984, 201 in 1986, and another 201 in 1987, have In the recent past been put into effect. Even with these increases the average rent, at about FMG 14,000, is still 2- 3 times less than market rents for housing of the same quality. The Incomes of SEIMAD housing occupants in Antananarivo is more or less twice the average income for the city, according to a January 1985 survey, with the result that only about 72 of income needs to be devoted to SEIMAb housing. In the Faritany capitals the civil servant housing stock of 3,300 units poses the same problems of very low rents and inadequate maintenance as in Antananarivo. The mismatch between rents and the resources of occupants is also apparent In the 781 SEIMAD units in the Faritany capitals, mainly located In Mahajanga and Antairanana, where occupants also spent about 7X of income on housing. 7. Although not a large part of the overall stock, civil servant and SEIMAD housing represents a supply of rental units that would be 11 As of October 31, 1989. - 162 - Annex 8.1 Page 3 of 5 difficult to replace quickly. A rehabilitation effort is therefore needed, and has already been 'aunched by SEIMAD following an Improvement in its financial situation. In view of the incomes of the present occupants, there is considerable scope for raising rents in public housing. This process could provide funds for maintenance and self-financing of new operations. It would also promote the turnover that ts desirable if these units are to fulfill their function of providing temporary accoamodation to the largest possible number of families. The transfer to a private agency (SEIMAD or another) of management responsibility for public housing, and new management methods (such as the creation by SEIMAD of a separate property mmangement affiliate) are two ways in which these objectives can be attained. B. Demand for Housing. Cost, and Affordability 8. Urban housing requirements are determined by the Increase in the numb r of households and the willingness of all households to spend in order to replace deficient housing and reduce the proliferation of one- or two- room dwellings. Considering growth rate trends and current estimates of household size, about 6,500 new dwellings would theoretically be needed annually in Antananarivo, and 1,300 in the secondary towns. It is difficult to estimate the needs associated with a reduction in overcrowding. Based on these calculations, ft would be desirable to triple current production in Antananarivo to 9,000 dwellings per year if the situation is to be significantly Improved. 9. The scope for adjusting the supply to meet thes" needs is very limited. Low levels of household income mean that effective demand Is far less than the theoretically calculated needs. According to the study of rural-urban linkages in Antananarivo, half the households in the Antansnarivo metropolitan area have incomes below FMG 92,000 per month. 2/ Average income Is around PEG 126,000 per month, but Is influenced by concentration at the upper endt 3X of households (senior civil servants, managers, expatriates) earn 42S of total Incomes. At the other end of the scale, 20% of households are below the threshold of absolute poverty (FMG 50,500 per month). 10. To estfimate housing Affordability in 1990, the results of the 1984 housing cost survey undertaken as part of the Greater Antanaarivo study have bean updated )/. The overall inflation rate was used for this purpose, even though it probably does not fully account for the reduction in the purchasing power of households over the past five years. The 21 Groupe Huitt/ARA, Etude des &changes economicues ville-camoamne dans la region d'Antananarivo, August 1989. 3I AURA/Groupe Huit/BCEOM, DeveloDoement urbain du Grand Antananarivo, 1985. - 163 - Annex 8.1 Page 4 of 5 construction categories are those used In the above study, 1.e., basic (1 self-built room, unfired bricks), limited (1 or 2 self-built rooms, unfired bricks with plaster, outdoor WC, simple wood finishings), intermediate (slightly larger, with better quality materials, utility connections, some subcontracting), and conventional (middle or high standard, subcontractor or firm). The results show that, at the present times a. th, bottom 30X of households, earning less than half the average income, cannot afford even a basic dwelling. Their only alternative is to rent dwellings whose total cost (less than FMG 1.5 million) assumes low land costs and a very rudimentary structure; b. the next 402, with Incomes from half of the average up to the average, have access to a limited dwelling, costing less than WM4 3.2 million; c-. 2O of households, those between the average and twice the average income, can afford a limited or Intermediate dwelling less than FMG 5.6 million in cost; d. an additional 52, between 2 and 3 times average income, can afford the intermediate standard at RMG 10 million or less; and e. only the top 52 of households can afford a conventional house. 11. Underlying the above typology is the economic crisis of the 1980s and the decline in the purchasing powsr. oLhouashol.&income,.ahich has resulted in a contraction in effective demand. Higher-standard dwellings have become even less affordaOle, the most striking regression being from middle-range to more basic accommodation through a reduction in livable space. 12. Based on the physical increase in the urbanized area and on surveys conducted by DAUH, the annual lncrease in the Antanaarivo housing stock is presently estimated at 3,000-3,500 units. This increase meeto only half of the Increase in population, the other half being absorbed through overcrowding of exlsting dwellings. What construction does take place Is frequently in the form of additions to existing houses, on family-owned plots registered as agricultural land, rented or purchased by the builder, or on public land with the agreement of the Fokontany. There has been no marked increase in squatter occupation of registered public lands in Antananarivo, although a new settlement is presently growing on the hillsides in Antananarivo-Renivohitra. 13. Of the 3,000-3,500 new dwellings built each year, an estimated 150-350 represent investment In higher-standard housing in residential noighborhoods. SEIMAD in 1986 resumed housing promotion activities, but has faced marketing problems In its current projects (150 units at Ivato and 67 ha, and an additional 150 In Mahajanga and Toamasina) largely because of an unfavorable relationship between price (FMG 10 million per dvelling) and - 164 - Annex 8.1 Page 5 of 5 location. The construction bottlenecks are In part financial. They may also be traced to the limited capacity for on-site production of building materials. The major cause, however, is the shortage of land, derived both from unclear tenure status and from the absence of Infrastructure and facilities. 14. The role of informal z.hannels in building materials and In construction remains dominant, and Is tending to increase. Thus, the small constructiors firm and the independent artisan (tacheron) are becoming more informal as they carry out small-scale works, hire and fire unskilled labor, and seek out other jobs. It appears evident that the overall productivity of the building trades is declining, a process difficult to reverse without a new stimulus of demand and a steady growth In production. - 165 - Annex 8.2 CONTRACTUAL SAVINGS PVR HOUSING: ILLUSTRATIONS (PMG) ----------------EXAMPLES--- A B 1 D Monthly income 150,000 150,000 300,000 300,000 Initial deposit 150,000 10,000 300,000 300,000 Regular monthly deposit 25,000 35,000 50,000 75,000 Total savings (36 months, capital + interest) 1,218,000 1,627,000 2,435S,000 3,458,000 Loan (72 monthly payments) 1,175,000 1,520,000 2,350,000 3,215,000 Monthly loan repayment 24,000 31,000 48,000 66,000 Total savings + loan 2,393,000 3,147,000 4,785,000 6,673,000 Beased onu Return on ieposits: 92 Interest rate on loans 141 Interest cn loan equal to twice the interest paid on savings - 166 - Annex 8.2 Page 1 of 2 THE MIQL PROJECT: EVOLUTION AND CURRENT STATUS 1. The initial proposal to create a contractual savings agency for housing, to be called SMutuelle Inter-Epargnants pour l Logement' (MIEL), wes prepared in 1987 by the M.nistty of Public Works with asslstence from Canadian consultnts. Extensive discussions followed within the Government and with potentially interested foreign donors. In May 1988, the Government endorsed the proposal to create a private non-profit financial Institution designed to attract private savings ad to make loaso for the purchase of bousing. It also decided that NIEL would have primary responsibility In this regard, and contli,vled the promotion efforts with domestic and foreign financial institutions. 2. The MIEL project ls based on the mutual-credit concept of organizing regular savings deposits, initially in this case by groups of 10 savers. Following a prescribed period of savings, loans for housing would be made. All savers in a group would have the same initial savings capacity, deposit schedule, and timing of house construction, which would help ensure a regular cash flow for the entity. The flnncaing plan provided for two standard savings plans of 42 and 44 months each, which during the initial phase could be reduced to 10 months with a larger down payment. Monthly savings deposits would be FMG 9,000 or 15,000, out of monthly incomes varying between 1MG 60,000 and 100,000. The savings would not veceive interest, but would result in a housing loan equal' to 5 times the accumulated savings, payable over 12.5 years at either 7S or 81 depending on the inLtial savings plan. For FMG 400,000 In Initial savings, a loan of NMG 2,000,000 could therefore be-granted. Of this amount,r-HIEL was to obtain 902 from external borrowing at an average rate of 42. 3. In addition to the savings groups, the MIEL project essentially depended on external borrowing at concessional interest. Total borrowing requirements over the first 7 years were estimated at FRG 21 billion (US$13.5 million), for savings averaging about FMG 2 billion (US$1.3 million). On this basis, through Income from investments and loan margins, the system was to become self-sustaining after 20 years. In 1989 several measures designed to make the program more attractive were introduced or modifleds the number of savings plans was Increased to 7; 4.5Z interest would be paid on savings; the minmum savings group was reduced to 5; and the celling on housing loans was raised to FMG 10 million. At the same tie the Interest rate on housing loans went to 102, and total external financing requirements to FMG 33 billion (US$21.2 million). In its most recent phase a MIEL promotion company (SP MIEL) has been created, with the objectives of Opromoting mutual associations for housing" and Amanaging the resources of mutual associations and foreign lenders.* 4. Despite these efforts, the MIEL project has not moved forward at tho expectod pace. Several problams have not been resolved. The most important is the very heavy reliance on concessional foreign borrowing to make the project viable, borrowing of a type which has beea quite rare for 15 yars. The abseonce of such funds first resulted in inactivity, then in - 167 - Annex 8.3 Page 2 of 2 some skepticism among those responsible for axplaining the project to poten*aWl savers and generating business. Furthermore, in view of the videly-known scarcity of buildable land, there was much uncertainty about the prospect of actually getting houses built. Finally, without experience NIEL has found lt difficult to organize mutual savings associations among groups (firm employees, civil servants, Fokontany residents) which often have their own housing offorts underway. 5. Monetary authorities also had reservations about a new agency without adequate risk coverage, drawing upon significant external financing in foreign exchange. Foreign donors in turn tended to regard MIEL as an instrument for introducing another housing subsidy, indirect in this case, in the form of lowered borrowing costs for savers. Replicability would not be assured since the Malagasy Government would not by itself be able to continue the program. Experience with such subsidies shows in any case that they are typically captured by the better-off families, those able to mobilize savings rapidly. The recent evolution of MIEL provisions, toward a greater number of savings plans and loans of FMG 400,000 and more, appears to strengthen this tendency. In summary, MIEL up to the present has functioned as a temporary source of funds, without adequate links to the Malagasy financial sector or assurance that Its social objectives can be achieved. IBRD 2003R .-rf ,Abranana MADAGAS:~~~~AR -vm -mo NO" ditim '\p f *O >2 NT IRANA 44 joe,tw 2 s E ¢ Nt;A 4"WNP, 0 Tp> S ) i J |Ih 4 \ FT"+ { ~~TOAMASIN . -'" "' ~~~~~~A 'Ai4oz ' ' mhbi '/ ° rv \_W w_BMF7 _ t j \ s"'S 'ANAN hdl : \> <~~ r o@; AlwojAntsi_ih . \ r Anira/+ / a t s ~~~~~nicr -~~~~~~~~~~~~~~~~~~~~~~~~OTR T Sp VL I FY 2w/ rA °0-~~~~~~~ f, yMononjory~~~~~~~~IIL . } f IANARgt$~~~~~~~~OlAmw tANANI CAPON" ft I I . /, ( ~~~~~~%* U /Aimssf Mo oNZNl ~~~ ""_ , An~~~~X_ i 4 { t f < ,' H1~~~~AIAGASt&AR d"tf_-~~~~~~~~ iS iifo S~~~~~~~~SOUTH w_a ' / ~~~~~~~~~~~~~~~~~~~AFItICAI iM RJNE 1990