Morocco - Housing Sector Strategy

Document of The World Bank FOR OFFICIAL USE ONLY Report No.13930-MOR KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY September 7, 19951 IWO Infrastructure Operations Division Maghreb and Iran Department Middle East and North Africa Regional Office 1Tis document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its comments may not otherwise be disclosed without World Bank authorization I1 Revised report (prior version issued March 20, 1995) which takes into account Government commnents of July 27, 1995. CURRENCY EOUIVALENTS US$1 = DH 9.0 ABBREVIATIONS AND ACRONYMS ANHI Public Land Developer (Agence Nationale pour l'Habitat Insalubre) CDG Saving and Public Funds Management Corporation (Caisse de D6p6t et de Gestion) CDVM Security and Exchange Commission (Conseil D6ontologique des Valeurs Mobili6res) CGI Housing Company (Compagnie G6n6rale Immobilibre) CIH Government Mortgage Lender (Cr6dit Immobilier et Hotelier) COS Floor/area ratio (Coefficient d'Occupation des Sols) ERAC Public Land Developer (Etablissement R6gional d'Am6nagement et de Construction) INAU Public Research Center (Institut National d'Am6nagement et d'Urbanisme) MH Moroccan Housing PA Land Use Plan (Plan d'Am6nagement) SDAU Urban Development Master Plans (Sch6ma Directeur d'Am6nagement et d'Urbanisme) SNEC Public Land Developer (Soci6t6 Nationale d'Equipement et de Construction) ZAC Comprehensive Development area (Zone d'Am6nagement Concertd) KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY 1995 CONTENTS EXECUTIVE SUMMARY ..................................... I. INTRODUCTION ANDBACKGROUND........................ 1 II. ANALYSIS OF SECTOR ISSUES ............................ 6 A. Constraints on Supply ................................. 6 B. Constraints onDemand ................................ 11 M . HOUSING STRATEGY ................................... 15 A . C ontext . . . .. . . . . .. . . .. . .. . .. . .. . . . .. .. . . . . . . . . .. . 15 B. Policy Recommendations ............................... 19 C. Low-Cost Housing Prospects Under the Above Policies ............. 27 D. Benefits from Recommended Policies ........................ 28 E. Institutional Recommendations ............................ 31 F. Action Plan to Carry Out the Reform ........................ 32 IV. BANK ASSISTANCE ..................................... 33 TABLES Table 1: The Urban Market Solutions: Supply and demand ................. 4 Table 2: Land Price Trends Average Price per M' by Year and Mode of Purchase (inDH) ...............' .......... 5 Table 3: Estimated Housing Sector Subsidies in 1992 ..................... 14 Table 4: Summary of Proposed Measures ............................ 17 Table 5: Housing Sector Prospects ................................ 30 GRAPHIC Lot supply in Morocco for the nextl5years........................... 31 CONTENTS (continuation) ANNEXES A. Complementary tables B. Housing sector strategy memoir (prepared by a Moroccan interministerial committee and the Bank mission) C. Report on the impact of urban planning and regulations on land use efficiency, by Mr. Francesco Bandarin D. Land price trends and instruments of land management, by Mr. Vincent Renard E. Housing sector strategy: Visible and less visible subsidies, by Mr. David Sewell F. Bibliography This report was prepared mainly on the basis of a mission which worked with a Moroccan interministerial committee in late October/early November 1994. It conveys primarily Bank views, but most of the policy analysis and recommendations are those of memoir (Annex B of the Report) jointly prepared by that interministerial committee and the Bank task team during the mission. The above mission was composed of Ms. Maryse Gautier, and Messrs. Julio Linares (task manager), Henri Beenhakker, Robert Buckley (TWURD), peer reviewer, Francesco Bandarin, Vincent Renard, and Frangois Vaillancourt (consultants). Earlier contributions pertain to Messrs. David Sewell (EMTIE) and Claude Hovnanian (retired Bank staff). Ms. Michelle Detwiler edited the report, and Mrs. Jocelyne Simmonds produced it. Messrs. Robert Buckley, Alain Bertaud (TWURD) and Bertrand Renaud (FSD), peer reviewers, provided advice. Messrs. Daniel Ritchie, Director, MN1, Amir Al-Khafaji, Division Chief, MNIPI, share responsibility for report approval. The task team, moreover, acknowledges its intellectual debt to the task team of Bank Report: Housing: Enabling Markets to Work, led by Mr. Stephen Mayo; their approach underlies the present report. The Moroccan interministerial committee was composed of Messrs. Cherif Tahiri, Director, Department of Planning and Programs at the Ministry of Housing; Abdelhai Bousfiha, Director of Urban Planning and Architecture at the Directorate-General of Urban Planning, Architecture, and Land Development, Ministry of the Interior and Information; and Abdelghani Guezzar, responsible for real estate and business finance at the Ministry of Finance. In addition, the Bank task team highly appreciates Mr. Mohamed M'Barki's (SNEC's Director General) illuminating feedback during the preparation of this task, and Mr. Abdellah Lezham's (INAU) excellent research advice during the mission. KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY EXECUTIVE SUMMARY i. Background. Morocco's record of sustained growth over the past decade is well- known. So too is there an appreciation for its well-managed macroeconomic stability, as reflected in its low and stable rates of inflation. Its more recent steps to privatize, open up the economy to market forces, and broaden participation in governance are the initial elements of a potentially far-reaching reform of Moroccan society. As part of this reappraisal of their basic institutions, the Moroccans have recognized that it is essential that reforms in the way the housing sector functions be integrated into this broader reform process. As a result, in November 1994, a Moroccan interministerial committee and a Bank mission jointly prepared a note (Annex B) proposing comprehensive policies to improve the performance of the housing sector. This report details how this reform could take place, and how it could be integrated into the broader reform process. ii. Main Issues. While the housing sector's performance has lagged behind comparator countries, reform of the housing sector is not essential for this reason. Rather, it is needed because without reform the broader reform process itself is threatened. For example, with less than 50 percent of the population in cities, Morocco's demographic shift to a more urbanized society is in many respects still before it. But, the current land supply system, dominated by public suppliers who provide lots to families of all income levels, is unable to accommodate this kind of shift. While the current system has after a fashion coped with the shift so far, its functioning has not been without significant costs: urban land prices have increased at rates almost 10 percentage points above annual inflation for the past decade, reflecting the system's shortfall in incorporating land to urban use. At the same time, Morocco has fallen farther behind in the provision of basic infrastructure services. Already 20 percent of urban families either live with another family or in barracks, and another 20 percent live in illegal houses built on land with unacceptable health and safety standards. iii. Although essential to increase the welfare of Moroccans, greater market- orientation will accelerate the demographic shifts now in process, placing even more pressure on cities and making social inequalities more apparent. In very basic ways, the current urban land development system clearly cannot accommodate the increased demands it undoubtedly will face. iv. Besides its general shortcoming in not allowing cities to accommodate population growth, the current system fails to target benefits to the poor: about 40 percent of lots produced by public land developers are put back on the market by private intermediaries at substantial gain, reflecting the lack of transparency of subsidies in the current system. As reflected in the recent announcement of a government program to double both the annual supply of dwellings for the poor and housing production generally, Moroccan authorities have shown that they are keenly aware of the failings of the distributional system. But, with maintenance of the present system, sustained expansion of supply is unlikely. Without change, Morocco will be unable to improve the distributional performance of the housing sector, let alone double housing ii production in a few years. The risks of a significant expansion of nontargeted subsidies to address a problem created by a rigid public supply system are very real. Changing the public and private roles in land development and subsidy provision are prerequisite to the creation of a better functioning, more equitable system of urban development. v. The lack of transparency and limited competition that characterizes land development also characterizes financial sector functioning. Tight control mechanisms have sustained a large government presence in the financial system. But, the proximate results of this system have been financial shallowness, lack of competitiveness, and lack of incentives for saving. The ultimate results have been a less effective financial sector which is less able to allocate funds to high return investments. For instance, compared to a group of countries in the region with a similar level of GDP, Morocco had lower domestic savings and financial depth, and correspondingly provided much less mortgage credit. vi. Among other policies, high reserve requirements and obligatory placing of bank resources in treasury securities have constrained overall lending to the private sector. Further, the limit on lending rates crowds out many business and mortgage loans that could be made at higher rates. Finally, the Caisse de D6p6t et de Gestion's (CDG) allocation of most pension and social security reserves, effectively closes the kind of market-based long-term intermediation which might channel funds to mortgages. In this kind of financial environment, it is not surprising that housing finance is similarly constrained, reducing both household access to homeownership as well as the types of longer term debt instruments that could foster greater financialization of the economy. Hence, reforms of the housing finance system would not only improve housing affordability; they would contribute to deepening financial markets and strengthening the banking system. Market-based mortgage lending would increase the investment alternatives available to generators of long term funds, help diversify bank portfolios, and could help establish the long-term element of the yield curve. vii. While housing finance reforms could be a key complement to broader financial sector adjustment, they should be closely coordinated with and cannot precede such efforts. Similarly, increasing the supply of housing finance without first removing supply constraints on the housing sector would mostly result in increasing housing prices rather than the number of houses produced. Thus, important sequencing issues are involved in efforts to increase the supply of housing finance. In particular, an essential prerequisite is to focus on those measures which facilitate a greater private sector role in housing development so that the potential supply response does not. abort. When the development of a more responsive system is in place consideration should be given to integrating housing finance reforms into the overall financial liberalization program. viii. The Government of Morocco recognizes the pursuit of this program of reform will take some time. Hence, a matrix of reforms and their sequencing and likely time horizons are listed in the attached table. But, to summarize the approach the following strategic themes should be highlighted: ix. Changing Roles of the Private and Public Sectors. To increase the supply of houses especially for the lower-income population, and to increase the supply of urban land, the Government needs, first and foremost, to shift the development of urban lots to private developers. These developers are much better suited than the public ones to give customers iii what they want and to extract from those with means the total price of satisfying those wants. Public developers can easily indulge in providing customers what seems aesthetically and politically appealing, charging below market prices even to high-income buyers and, at best, minimizing costs rather than maximizing value. x. With urbanization pressures about to intensify Morocco has no room for waste and avoidable giveaways; every dirham spent should produce full value generally commanding full prices in free markets. that attract capable entrepreneurs. Under this approach, the Central ,Government would retain responsibility for formulating financial, urban, and housing sector policies, and provide transparent subsidies to those most in need. A key principle of such an approach is that the responsibility for urban planning and land management should eventually be transferred from the center to municipalities. Around these main themes, a supply, demand, and institutional strategy for 1995-2005 is proposed below (the attached table, providing a time frame, details the measures). xi. Addressing Supply Constraints. (a) Redefine the role of public developers, charging them with primary responsibility for developing land with main infrastructure, leaving to the private sector the role of developer and seller of lots.2 During the next few years, public developers would retain responsibility for developing lots only for the very poor. (b) Streamline the approval of urban plans and permits and revise urban standards downward, to reduce delays and uncertainties to developers and to contain the cost of land development. (c) Simplify and increase taxation of land property values, lower import duties on key construction materials and, to the extent possible, eliminate tax exemptions that under monopolistic markets benefit mostly land owners and developers rather than households. xii. Addressing Financial Constraints. (a) In concert with efforts to liberalize financial markets, the Government should foster the development of mortgage backed bonds issued by mortgage originators for placement with generators of long term funds; this would require Conseil D6ontologique des Valeurs Mobili6res' (CDVM) regulating the issuance and trading of such bonds. (b) Withdraw the Government's explicit guarantee on debt issued by government agencies to somewhat level the playing field among issuers of debt. (c) Clarify and enforce the right to rapidly foreclose on mortgage property so that recourse-based lending can take place. 2/ Public developers would fulfil their role through the following: on public or purchased land, selling tracts to private developers through competitive procedures; on land divided among few owners, agreeing with them on joint development (ZAC); and on land divided among many owners, administratively resolving to incorporate land to urban use (replotting). iv xiii. Housing the Poor. (a) Much greater reliance on clearly identified and budgeted subsidies should be used as the mechanism through which housing assistance for the poor should be provided. (b) In most countries, the existing housing stock rather than new production is used to address the housing needs of the poor. While Morocco's bidonvilles and significant share of housing in illegal settlements may require direct assistance (see the discussion above under the first supply side recommendation) attention should also be given to measures to improve the utilization of the existing stock. Rent control reduces this flexibility. xiv. Institutional Recommendations. (a) Transfer responsibility for urban planning and land management from the fragmented communes to the city wide "communaut6s", as such activities call for a city, not only a neighborhood focus. (b) Withdraw land development powers from the Urban Agencies, as such a role conflicts with their more important and appropriate one of supporting urban regulation. (c) For the longer term and within the context of state design, the placement of the entity in charge of urban management in the structure of the Government ought to be carefully considered. Ideally, to foster accountability, urban management should be vested in a department or agency dedicated exclusively to such a function. The Ministry of the Interior, with its several important, demanding functions, may not be the best place; nor is the Ministry of Housing as the mission of housing the poor may conflict with the more general one of regulation. A High Commissariat under the Prime Minister appears a promising alternative. xv. To sum up, the above approaches would integrate Morocco's housing supply system into the broader societal reform that is now beginning. They would also reduce housing costs, significantly increase the ratio of housing finance to total financing, and meet at least the housing demand resulting from population growth without increasing government expenditures. Bank assistance ought to underpin the above policies, and in the immediate future ought mainly include funding for increasing land supply serviced with primary infrastructure so that private developers can supply lots. Perhaps most basically, however, this reform program could make a significant contribution to unleashing the forces of the private sector in labor intensive and productive ways. V Summary of Proposed Measures Components Actions to take during Actions to take during 1995-1997 1998-2000 1. Streamline and operationalize - Accelerate review and approval - Introduce land development urban planning of urban documents. phases in urban planning - Introduce lower standards for documents. lots aimed at low income - Optimize operational urban households.' standards and revise regulations - Establish system for public to give more flexibility to the sector to launch ZAC and private sector.' replotting operations, and to - Combine land survey with take private land to recover development plan, to orient costs of, and benefit from, private sector decisions. urban expansion." 2. Develop information tools. - Start producing charts of market prices. 3. Let developers design - R6gies permit developers to infrastructure works. design and carry out works, subject to Regie standards. 4. Modify role of public land - Steer them towards providing - Launch ZAC and replotting developers, primary infrastructure and fully operations in the main cities.b developed lots only to the very poor.a b - Launch initial sales of tracts with primary infrastructure to private sector developers in the main cities.b - Promulgate requisite laws and plan for ZAC and replotting operations.' 5. Improve housing sector - Lower import duties for key - Simplify real estate taxes, taxation. construction materials.b expand the territorial base, and update such base in tandem with inflation. - Eliminate real estate tax exemptions and related nontransparent subsidies. 6. Deregulate rents. - Pass law allowing the rapid - Enact law streamlining rent eviction of tenants in arrears. control. - Sell public rental houses. 7. Liberalize interest rates. - Suppress limit on lending rates.d vi Components Actions to take during Actions to take during 1995-1997 1998-2000 8. Facilitate mortgage finance. - Issue CDVM's regulations on - Lower somewhat further the the issue of Mortgage bonds. percentage of reserves insurers - Withdraw government explicit must place in government guarantee on obligations of guaranteed obligations. public agencies.b - Extend rapid procedures to foreclose mortgages to all financial institutions. 9. Improve the institutional set - Transfer urban management up. responsibilities from the communes to the communaut6s. - Withdraw land development mission from urban agencies. a These measures require further study before, and some gradualism in, implementation. b Measures which best combine facility, convenience, and urgency, and thus require immediate action (early phase of 1995-1997 period). c Action which is the backbone of the proposed urban land management strategy. d Financial sector liberalization measure which is essential for mortgage markets to work. Other measures to liberalize money and capital markets should facilitate mortgage finance. KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY I. INTRODUCTION AND BACKGROUND 1. In spite of the Government's substantial progress in regulating urban development and eliminating shanty towns, the supply of adequate dwellings that comply with established urban standards falls far short of demand (complementary tables in Annex A). Furthermore, the formal private sector develops only a minor share of total urban land, the banking system devotes very little resources to housing finance, and the public sector assists only a part of the needy to obtain housing with a significant amount of help going to the well-off. Land management overwhelmingly constrains urban housing development, mainly through still cumbersome urban planning, and a lot supply system dominated by public sector agencies, system which by and large accommodated urban expansion needs during the 1980s, but now decays under diminished public land and increasing urban population. Housing in rural areas is generally thought to be less of a problem, with the availability of safe drinking water being the dominant interest in those areas. In late October-early November 1994, an interministerial committee of Moroccan officials and a Bank task team reviewed sector issues and proposed a comprehensive set of policies to resolve them, having documented their ideas in a jointly prepared memoir (Annex B). The most far reaching recommendation is that the Government empower the private sector to take over primary responsibility for developing urban land from the now dominant public land developers; for this, public land developers should phase out their present role as providers of urban lots for all classes of households (an activity which includes nontransparent subsidies and crowds out private sector developers), and assume instead the role of land supply facilitators by providing main infrastructure, putting land at the disposal of private developers, and retaining some land to develop lots only for the very poor. Other recommendations pertain to urban planning and regulation, taxation, and housing finance. This report, in addition to providing some necessary background, conveys Bank views on the analysis of issues and policy recommendations (mainly on the basis of the above memoir), and recommends some adjustments to the sector institutional set up and action plans to carry out the recommended policy and institutional changes. 2. Country Background. Morocco is a middle-size, low-middle-income country working its way out of macroeconomic adjustment into structural preparedness for growth. It has a population of about 26 million, growing at 2.3 percent per annum overall and 4.3 percent in urban areas. Morocco's gross domestic product has been growing at about 5 percent annually over the last few years, which allows for only minimal increases in its low per capita annual income of about US$1,000. Annual inflation is about 5 percent and the financial sector has liberalized to some extent, but the government budget has been chronically tight and the treasury still constrains the flow of financial resources to private sector financing. The size of public sector expenditures, over 25 percent of GDP, mav significantly restrict Morocco's growth. These expenditures plus the undeveloped financiaj sector combine to maintain Morocco's domestic savings rate at about 17 percent of GDP (the national savings rate is higher because 2 of transfers from Moroccans working abroad), compared to around 20 percent in developing countries of the Mediterranean basin and around 25 percent in high growth countries of Asia and Latin America. To accelerate growth and reduce poverty, Morocco will have to privatize public enterprises, which it has begun doing, further liberalize the financial sector to allow for an efficient capital market, and improve the availability of basic, especially primary education. Furthermore, it will have to substantially reform key sectors that cut across or could support those priorities, so that government regulation facilitates private action, and markets work efficiently and equitably. Housing is one of these key sectors. 3. The Housing Sector in the Macroeconomy. Upgrading the efficiency of the housing sector is of primary importance for the economic development of Morocco, for four main reasons. First, addressing housing problems requires addressing the scarcity of urban land, a strategic resource that underlies the cost structure of industry, the total cost of urban growth, and potential mechanisms for regressively redistributing wealth through, among others, non recovery by society of the land values added by urban expansion (para. 8). Second, this sector, which accounts for in the order of 5 percent of GDP and a much larger share of total investment, is a significant part of the economy. In a housing sector with constrained supply, as is the case in Morocco (para. 10), enhancing the supply response could lead to substantial increases in welfare for the same or lower level of housing expenditure, by reducing inflated prices of houses and increasing the number of houses supplied. Third, housing expenditures loom large in the budget of the poor; they are at least one third of their income, and, therefore, reducing housing prices or targeting subsidies better could substantially improve social conditions. Fourth, from the financial perspective, housing finance arrangements could deepen financial markets, by, for example, adding sound mortgage-based obligations to investor portfolios, and establishing an element of the yield curve. Thus, effectively addressing housing issues is bound to significantly improve the efficiency of several real and financial markets and the functioning of the economic system. 4. Sector Institutions. The public sector institutional set up is somewhat complex, involving diverse types of public agencies which interact mainly in the supply of urban land. On the central level, the Ministry of Housing and the Ministry of the Interior, in effect, share most of the responsibility for the housing sector. The Ministry of Housing formulates housing sector policies and oversees several public land developers which dominate the supply of urban lots country-wide. The Ministry of the Interior formulates urban land policies, and in particular approves urban master and zoning plans, and even clears small real estate developments in the municipalities it is responsible for overseeing. On the operating level, two national agencies, Soci6t6 Nationale d'Equipement et de Construction (SNEC) and Agence Nationale de Lutte contre l'Habitat Insalubre (ANHI), and seven regional agencies called ERACs, are responsible for providing mainly lots (and some houses) to all classes of households, cross-subsidizing the poorer with higher charges for the well-to-do, a system called "p6r6quation" (cross- subsidization). The Government also has a housing developer for Casablanca, Attacharouk, and the publicly owned Caisse de D6p6ts et de Gestion (CDG) has real estate affiliates dedicated to the promotion and construction of middle-income housing. Municipalities, for their part, are primarily responsible for approving land developments and housing construction, but subject to the approval of the governors and even of the Ministry of the Interior for all projects of significance. An existing Urban Agency under the Ministry of the Interior in Casablanca (and 3 future similar agencies being created for other major cities) is responsible for reviewing urban plans and developments in that city, and also for developing land in competition with the above public agencies and the private sector. A government-owned Bank, Cr6dit Immobilier et H6telier (CIH), is charged with providing housing financing to low middle-income households. 5. The private sector, as characterized by consultant surveys, includes: formal land developers, composed of a small number of professionals and a myriad of amateurs which enter the business because they own land susceptible to urbanization, a very scarce, expensive resource; individuals building their own house in a purchased lot, the predominant practice in Morocco; construction companies of varied sizes and capabilities; informal land developers; private banks with minimal mortgage portfolios; and professionals, of which independent architects must endorse all significant land developments and housing projects. 6. The Market Solutions. As noted above, the supply of dwellings fails to satisfactorily meet demand. During the 1980s, the growth of urban population alone has called for the annual production of around 89,000 dwellings, whereas suppliers have delivered only 80,000, including some 27,000 unauthorized, of which around 4,500 are of shanty town standard (Table I and Annex A-1). Public land developers, with administered standards and prices, account for 60 percent of the formal supply of total lots, leaving a minor share of the total market to be satisfied on free market terms. Consultants to the ministry of Housing (Team Maroc study) have shown that about 40 percent of lots produced by those public developers are put back on the market by private intermediaries at substantial gain, which more reflects nontransparent subsidies and excessive standards than the solution of a social problem. Morocco has a ratio of house price to annual family income of 6.7 -- the highest among countries of similar GDP per capita -- compared with an average of 4.1 for these countries and 3.4 for Jordan (United Nations-World Bank Housing Indicators Program -- Annex A-7). In the formal sector, a 70 m2 dwelling generally envisaged for low-income families now costs over DH 200,000 (US$22,000 equivalent -- as per practices of public developers and financiers, such price does not include the full market price of the lot) which shows both excessive standards and high resource costs. Developed urban land commands extremely high prices which often account for over 40 percent of the cost of a house, even though lots accommodate more than one dwelling and most land data does not record full market prices. Non-authorized dwellings in the informal market cost less than half those of the formal market, reflecting partly adjustment to needs and capacity to pay of the population, and partly poor urban standards that might be unacceptable from a health and safety standpoint. These conditions call for, more than overcoming limitations of financial resources for housing, reorganizing the system of production of lots and houses to appropriately meet the needs of Moroccans. 4 Table 1: The Urban Market Solutions: Supply and demand Annual average for 1981-1990. Low income Middle-income High income households households households TOTAL L07S Public Agencies 9,000 8,600 4,700 23,500 Private formal NA NA 7,300 7,300 suppliers Informal Market 3,500 3,500 NA 7,000 Bidonville solution 4,500 NA NA 4,500 TOTAL 42,300 HOUSES Public Agencies 300 3,900 NA 6,900 Private formal NA 8,100 33,800 41,900 suppliers Informal Market 13,700 13,000 NA 26,700 Bidonville solution 4,500 NA NA 4,500 TOTAL 80,000 Annual needs of urban dwellings based on population growth 89,000 Prepared on the basis of Annex A-1. 7. The Scarcity of Urban Land. The dwellings supply system seriously fails in supplying urban land, the prices of which show an alarming, increasing trend. According to the above Housing Indicators, for example, Rabat-Sale's land development multiplier (ratio between price of developed to undeveloped land in the area) is 5.5, about 40 percent in excess of Tunisia and Amman, excess which reflects Morocco's more difficult incorporation of land to urban use. A 1991 survey by consultants to CIH shows that average lot prices per m' in 1989 were DH 815 equivalent for Morocco, compared with DH 690 equivalent, for 1993, in the Governorates of Tunis and Ben Arous in Tunisia, a country which also has a constrained supply of urban land. Morocco's land prices, moreover, have escalated substantially in the recent past: in accordance with that survey, prices per ml of housing lots sold by persons who bought from a public developer (sales which reflect market conditions) grew at an annual rate of 21 percent during 1982 - 1989, far in excess of annual inflation of around 6 percent (table 2). A growth rate below inflation for prices of lots sold by public developers reflects not market conditions, but increasing, nontransparent subsidies for the buyers benefitted with those lots and the high profits realized from resales. Urban land escalates not only in major, but also in small cities: A survey by the Urban Agency of Casablanca shows that land costs per m2 of housing floor area have escalated 7.5 percent annually in real terms during 1984-1990 in major cities and 9.7 percent in smaller cities, on balance about doubling the price escalation of other housing costs (Annex A-2). Commercial and industrial land prices underscore the alarming trend: data from CIH's reports show that prices of land suitable for commercial buildings in Rabat and Casablanca increased 250 percent during 1986-1991, compared to a 25 percent increase in the consumer price index; the study Foreign Investment in Morocco (carried out by Price Waterhouse under IFC/MIGA sponsorship) shows that Casablanca's industrial land per m2 cost is more than double 5 that of Tunis, and Tangiers' more than double those of Marseille and Valencia. Urban land scarcity may even worsen, as Morocco, with 48 percent urban population, is far away from its potential urbanization peak and around 10 percentage points below countries such as Jordan, Tunisia and Turkey; Morocco's urbanization will most probably catch up, and thus put further pressure on urban land. Land scarcity, thus, appears to be a major problem for the housing sector and the overall economy. Table 2: Land Price Trends Average Price per M2 by Year and Mode of Purchase (in DH) Rate of increase in annual prices 1982 1984 1986 1989 1982/1989 Government transfer 189 197 303 238 3.3% Purchase from government transferee 242 516 962 935 21.0% Direct purchase from private individual 510 849 1028 1205 13.0% Purchase through middleman from private 600 1159 1506 1368 12.5% individual Direct purchase from developer 312 648 788 716 11.3% Purchase through middleman from developer 600 420 538 1018 7.8% Overall 328 390 605 815 13.9% Source: ClH's Quantitative Survey, 1991. 8. Serving the Poor. Although the Moroccan authorities strive to assist modest income households, the public supply system of lots, in particular, falls short of meeting the needs of the poor (population at the third lowest level of consumption). For 1981-1990, the Bank estimates that such a system provided some form of housing assistance to the equivalent of less than one third of the annual increase in new poor households (Annex A-3). Urbanized lots are so scarce that the privileged, benefited families manage to construct their dwelling mainly by auctioning the right to build a second and even a third floor on their lot; but the majority of the poor must share dwellings, live in shanty towns, or at best live in stable dwellings on land that does not meet minimal urban standards. In 1992 and 1993 the Ministry of Housing and related public developers managed to target about 90 percent of the 53,000 produced lots to those below Morocco's mean level of consumption. Still, such achievement appears to fall short of assisting about one half of the increase in new households at the third lowest level of consumption; it may also be unsustainable in the medium term, as maintaining such programs aimed at families of modest income would call for increasing government subsidies or availability of public land. 9. Government Plans. Following royal directives, the Moroccan authorities are embarking on a drive to increase the annual supply of primarily urban dwellings from around 100,000 to around 200,000 country wide. This, at the minimum, calls for doubling what public developers now produce, if existing production shares are maintained. A preliminary program already exists for some 25,000 houses designed according to traditional lot standards of about 100 m2 and construction of about 70 M2. Expected budget resources for this effort appear scanty--some DH 250 million of transfers for public land developers and DH 550 million of 6 interest subsidies, following practices of the recent past; but it is apparent that reorganization of supply, more than budget transfers, is what the Government plans require. As the Government and the Bank jointly developed the policies proposed here when preliminary plans were just put forward, those plans will have to be adjusted to the present proposed policies. Under present supply constraints, as noted below, Morocco would be unable to improve the present performance of the housing sector, let alone double housing production in a few years. H. ANALYSIS OF SECTOR ISSUES A. Constraints on Supply 10. As noted in the jointly prepared memoir, deficiencies in the cycle of housing supply, from urban land regulations to the delivery system of urban lots, largely account for the weak performance of the housing sector, including the alarming increasing trend of urban land prices. The following paragraphs analyze housing supply constraints (detailed analysis of urban planning practices, and land markets are presented in Annexes C and D, respectively). 11. Urban development plans. Slow and uncertain management of urban planning regulations, more than deficiencies of urban plans, constrain the production of urban land and slow the cycle of housing supply. The linkages between the urban planning instruments, Sch6mas Directeurs d'Am6nagement et d'Urbanisme (SDAUs, or Urban Development Master Plans) and the Plans d'Am6nagement (PAs, or Land Use Plans), defined by the law of 1992 seem to be satisfactory. The provisions of that law, moreover, call for processing urban regulatory decisions compatible with efficient urban management once those provisions become operational. The output of those plans meets requirements. At this time, 17 SDAUs covering Morocco's major metropolitan areas and 217 PAs have been or are about to be approved. Yet, delays in preparing and approving the documents inhibit lot production and urban development. This situation makes land owners' and real estate developers' returns uncertain, and destabilizes the market. From a quantitative standpoint, the land use plans generally provide for an adequate future supply of lots for urban development. Even so, from the standpoint of implementation, the above plans lack timetables for developing and financing land supply projects and for carrying out land improvements, a situation which could lead to spatial dispersion of development and under-utilization of infrastructure. On-going projects, mainly those undertaken by public developers, often are developed as a function of land availability, independently of their location relative to the land use plan. In implementing land use plans, market analysis and land survey components, which would optimize development options, are often inadequate. 12. Urban planning standards. The standards specified by the urban development plans seem to meet international specifications routinely applied to regulated urban development projects, mainly in developed countries (Annex C). These demanding standards could constrain especially the development of low-cost lots for low-income families, but an analysis of various public land developers' projects shows that the land use ratio was even lower than permitted by such standards and warranted under the current land shortage. The land use ratio varies from 35 percent to 45 percent, while statistics on public and low-cost housing projects in other countries show that land use ratios in the 55 - 65 percent range can be attained. Similarly, while housing densities found in unauthorized housing developments do not serve as a reference, they suggest a clear-cut need for more efficient land use, especially from the part of public land 7 developers. Still, differentiated standards for public housing projects and for slum upgrading have sometimes been adopted in practice as permitted by present laws. 13. Functional urban planning. Urban development documentation (SDAUs and PAs) have little relationship with the location of urban development projects. To a large extent, the availability of land, primarily public land reserves which are being depleted, and large, privately-owned tracts, dictates the location of new developments. The PA, therefore, does not seem to be fulfilling its role, which is to serve as a framework for the production of building lots. Linking urban plans to land development through operational and regulatory means to bring land into the market appears essential. 14. The Procedures for Approving Urban Documents and Permits. Furthermore, the procedures for approving urban planning documents and issuing land development and construction permits are complex and lengthy (details in Annex A-4 to 7). The Government, by decree, approves SDAUs which would guide urban development at the horizon of 25 years, after such plans, designed by consultants under the Ministry of Interior, have been agreed among the ministries of Public Works, Finance and Agriculture, plus the respective municipalities and city-wide governments. The Government, by decree, also approves land use plans, which are followed to authorize land developments. The Department of Urban Planning of the Ministry of Interior or the Urban Agency, with the assistance of consultants, prepares those plans, and revises them three times to accommodate sequential comments by municipalities, participants in a public inquiry and the above ministries, before that decree is ready for passage. This process has proved extremely slow. Land development permits, for their part, require approval by the Ministry of Interior when they accommodate more than 100 dwellings, and agreement between the local representative of that ministry and the municipality below that limit. Municipalities issue building permits, with clearance from the local representative of the Ministry of Interior. By law, Government inaction on development and building permits implies approval, but minor comments may lengthen the approval process indefinitely. Because of the absence of a common form for these permits, moreover, authorization practices are unclear, lack consistency across cities, and make the business of developers uncertain. 15. Land ownership and registration. Land ownership and registration, for their part, appear to be no major problems. The army and religious organizations have substantial land out of the market, but cities grow with reasonable efficiency around such land. About 30 percent of urban land countrywide (90 percent in small areas of the north) is unregistered, but even such land can be sold, and, when registering the sale after a slow process, be eligible for mortgage financing. 16. Provision of Infrastructure by the Local Sector. The local sector contributes little to the urbanization of land. Municipalities and their agencies (utilities, called R6gies) are passive in expanding primary infrastructure, weak in recovering installation costs when expanding service networks, and constraining as they demand that they study those works required by public and private developers. This may have facilitated the dominance of public developers in furnishing urban lots (para. 18). A long process of local sector reform and upgrading of intergovernmental arrangements will be necessary before the local sector becomes pro-active and financially minded in the development of land for urban use. Public land 8 developers may have to fulfill the function of providing main urban infrastructure, including some off-site, at least through this decade. 17. Managing the land market. Under prevailing conditions, moreover, the operation of the land market lacks transparency, and thus compounds the above factors for increasing land prices. While information on land prices is available locally, there is no system for consistently monitoring market conditions. Extensive under-reporting (for tax-related reasons) of prices declared when recording land transactions contributes to lack of transparency. But publication of some transactions at very high prices could contribute to disseminating the notion that prices ought to be higher than warranted by real market conditions. 18. The Public Supply of Urban Lots. As noted above, the public system of providing serviced lots has been unable to adequately expand land supply and serve the needs of the poor, and this inability is likely to become even more pronounced as time goes by. Typically, public land developers, who use land inefficiently (para. 12), develop government land or land obtained under some threat of expropriation and, by and large covering costs, sell lots at about a 15 percent markdown (lowest markdown mentioned by public officials) from market to the well-to-do, and a more than 50 percent markdown to the poor; this system is called "p6r6quation" (cross-subsidization). The scarcity of land with primary infrastructure, the gradual depletion of government land, and the weakness of small land holders has made this apparently inconsistent operation possible. 19. The above system fulfilled some government objectives in the conditions of the 1980s, but is proving unsuitable to foster land and housing development in the conditions of the 1990s and beyond. During the 1980s, the Government needed to expand urban land to house a growing urban population, but without using budget resources which an untamed deficit and an international debt overhang kept out of reach. Public land in the urban fringes, moreover, was much more available than it is today. Under these conditions, it would have been difficult for the authorities to have developed a system of land supply other than the present cross- subsidization. But with public land substantially diminished and landholders better resisting the threat of expropriation, the ratio of lots for higher income families to lots for low-income families has almost doubled for the operations to be financially viable. The system, thus, serves less and less the needs of the poor. With this, and 40 percent of lots purchased from public developers' being resold at a substantial profit, the system of cross-subsidization is becoming a system of benefitting some middle and high-income families with low-cost housing and sources of cash. With the natural incentives of public enterprises, it is unrealistic to expect that public developers consistently extract full market prices for lots from higher income families and subsidize only the very poor. 20. The present cross-subsidization system of land supply not only falters in channeling targeted subsidies to the poor, but appears inherently inadequate to expand land supply. The review of several cities (Rabat-Sal6, Casablanca, and Khemisset -- by Bank staff and consultants -- Annex A-8) shows that public developers retail part of the total needed serviced lots at prices below those of the private formal market, and freeze private development in their areas of future operations while they negotiate or expropriate the necessary raw land. As a result, raw land prices skyrocket in areas proximate to local services and free from the intervention of public developers, giving a windfall profit to landholders in such areas. Formal 9 and informal private land developers, often the landowners, thus are able to retail lots at what the market would bear, deriving large operating margins. Hence three separate land markets coexist in each city: the rationed market of public developers, which retails mainly lots with ambitious urban standards to low, middle and high-income households at below market prices; the free formal market, which retails lots and houses to people with means at the full price commanded by scarce urban land; and the large informal, nonregular market, which supplies mostly lots developed below those standards to people of low and low-middle income. A segment of the poor qualifies for no market, and develops shanty towns. The system, thus, has been largely inequitable. Besides leaving aside most of the poor, it includes the above- mentioned nontransparent subsidies embedded in the below- market price of lots for the well-to- do and windfall profits for landholders not threatened by expropriation; the total population and weak landholders have ultimately paid for those subsidies, through depletion of public land and being targeted to sell under some threat of expropriation, respectively. 21. Still, even if the Government could command much more raw land for development by public agencies, the failure of the supply system could not be remedied by continuing to crowd out private sector supply. Further increasing public supply of lots does not appear viable as public land is rare and landholders are now much more effectively resisting expropriation and demanding market prices. Even if extending public supply of lots were viable, such a supply would neither be as responsive to market signals as the private sector's nor be subject to a hard budget constraint, and thus, more likely than not, would give less social welfare per dirha.m spent. The observed tendency of public developers to use land less intensively than allowed by somewhat ambitious urban standards illustrates this point (para. 12). Under these conditions, expanding public supply would call for expanded recourse to central budget funds and still fall short of meeting the needs of the poor. In its time, the public supply system of lots permitted the provision of dwellings for the poor with controlled burden on budget cash flows, and afforded substantial cost recovery on land improvements. But such a system, as noted, has lived well beyond this limited, declining usefulness. To improve sector performance, the Government must increase private participation in the provision of urban lots (para. 35) and find other means to recover the costs of urban growth. 22. Land taxes. In spite of substantial improvement in tax administration recently, taxes assessed on land and real property (urban tax, council tax, and tax on nonconstructed land) have much room to improve from the land management perspective. This is particularly the case for the tax on undeveloped lots, which was reinstated to serve as an incentive to use or sell urban land, but which plays only a marginal role because of its low rate and its limited scope of application. There still is no form of taxation on land located outside urban perimeters. Hence, it costs nothing to store such land from a tax standpoint, even though its location in urban fringes and the proximity of infrastructure substantially increase its market value. Also, the tax to be collected by local governments to recover the value added to land by public services is not applied. Finally, the authorities adjust the tax base only every five years upon reassessments (they do not make intermediate adjustments in tandem with inflation), which maintains the tax base below market values, and distorts taxes among taxpayers and across regions. 23. Constraints on the supply of rental housing. The regulatory framework for rental units also contributes to restrict the supply of housing. Under the current law, delinquent 10 tenants can be evicted only by following the customary civil procedure, which is cumbersome and time consuming. Moreover, this law specifies that rents can be increased only every three years, and only after filing a petition with the court, if the tenant objects. As a result, investors are reluctant to invest in rental property or to rent dwellings that were built for eventual sale. Hence, thousands of urban dwellings now stand vacant because their owners believe that renting them to tenants who may not pay or vacate could prevent them from recouping their investment. A bill now before Parliament proposes to solve this problem by enacting more efficient procedures for evicting delinquent tenants. However, it maintains the legal procedure for raising rents only every three years. It would appear that this escalation provision will remain in effect until such time as streamlined eviction procedures and land market liberalization have considerably increased the supply of housing, particularly for low-income households. Moreover, Sophal/CIFM's (an affiliate of the CDG) managing about 54,000 public housing units that are rented at rates substantially below market underscores the adverse effects of restricting the free contracting of rentals, and of having the public sector as landlord. 24. Import duties. Another factor which constrains the supply of housing is import duties. For example, the ad valorem tariff on cement is 25 percent, which must result in a commensurate increase in the price of this commodity. Since there are several cement factories in Morocco, and cement should be produced efficiently in all countries of intermediate size, this import duty and the attendant high price of cement results either in continuing inefficient production of cement, or in generating monopolistic profits, at the expense of people who need housing. Similarly, the import tariff on concrete reinforcing rods is 52.5 percent. According to the above Team Maroc Study, from which this information was abstracted, the only factory of any significant size which makes reinforcing rods is operating far below capacity. This implies that the import duty protects inefficient local production and also increases the price paid by the end users. 25. Other fiscal measures. Direct or indirect fiscal subsidies incorporated in tax laws, in the order of DH 2 billion (Annex A-9), have either failed to increase supply, or have resulted in higher demand from groups other than low-income. Because these potential buyers face a monopolistic supply from real estate developers, the tax exemption on real estate profits (Morocco Housing Sector Strategy: Overt and Hidden Subsidies -- Annex E) results in higher profits for private real estate developers rather than an increase in the supply of building lots and housing. In addition, the deductibility of mortgage interest payments from gross income is more advantageous to wealthier homeowners than to lower-income groups. In Morocco, as in all developing countries, the tax deduction for mortgage interest is far more regressive than in developed countries. In developing countries, low-income groups (most of the population) do not have ready access to the mortgage market, and that deduction, which benefits the affluent and middle classes, reduces the revenues available for direct assistance to the poor. In developed countries (where most of the population is roughly middle class and has access to the mortgage market), this deduction mainly results in overconsumption of housing, which benefits the real estate sector. 26. The above supply constraints, on balance, affect mostly land, and result mainly in increasing land prices to the majority of households. The slow approval of urban documents and permits, the limited expansion of private infrastructure, the freezing of markets in zones of intervention by public developers, and minimal land taxes, in particular, combine to boost prices 11 of raw and developed urban land in the free, non-administered segments of the market. Often times, constrained land supply, like in Morocco, results in gradual growth and compact, dense cities, and, thus, in low cost of urban growth; but this does not appear to be the case for Morocco. The inefficient use of land by public developers who dominate land supply, of necessity, raises the cost of urban growth. Thus, overall land management, for now, gives Morocco some of the worse of two, almost incompatible worlds: the high prices of land associated with constrained land markets, and higher than warranted costs of urban growth, generally associated with unconstrained urban land markets, which Morocco is not. Regulation, to some extent, but mainly public supply of lots driven by nonmarket stimuli, seem to make compatible those generally incompatible results. The positive aspect of this is that Morocco could be able to pursue concurrently both policies that reduce the scarcity of land, and policies that contain the cost of urban growth. B. Constraints on Demand 27. Lack of Mortgage Financing. It is generally recognized that a lack of long-term financing reduces demand for housing, particularly from modest-income groups. Only some 10 percent of households finance their home purchases through the financial system; the amount of mortgage loans represents about 7 percent of the amount of total bank loans in Rabat, compared to about 20 percent in Santiago (Chile) and 44 percent in Washington, D.C. Private banks allocate a minimal part of their resources to housing, providing primarily three-year loans for housing development to enterprises and a limited number of ten-year mortgages for houses of a minimum cost of DH 700,000 aimed at families with income above DH 20,000 per month. CIH finances houses mostly in the DH 200,000-300,000 range at below-market (5 - 7 percent) interest rates, and obtains interest rebates from the Government to keep interest earned at allowed maximum rates for long-term loans; these houses are generally aimed at income levels of less than DH 5,000 per month. For most middle-income families, access to financing, even on market terms, is nonexistent (Annexes A-10 and 11). 28. The Financial Sector Environment. As conveyed by Government-Bank dialogue (Bank note of February 1995, aide-m6moire of November 1994 and Bank report no. P-6663- MOR of June 29, 1995 on a Financial Markets Development Loan), the level of financial sector development is not conducive to developing long-term financing, particularly for housing. The weekly auctions for selected maturities of Government securities are thin and nonhomogeneous, and thus fail to provide a maturity schedule of interest rates. The lack of a secondary market for treasury securities requires that Bank Al-Maghrib manage credit levels mainly through changing banks' reserve requirements, which forces banks to adjust lending on short notice inconveniencing both them and their borrowers. The very limited capacity of exporters to keep foreign exchange and other related restrictions virtually preclude exchange risk management. Regulations are not in place for, among others, the issue and sale of private debt through the exchanges. With these conditions, the interbank money market is opaque, the treasury debt market nonfunctional, and the private debt market nonexistent (no debentures have ever been traded in the Stock Exchange and trade in stocks is shallow); savers and investors cannot assess term, currency, source, and risk structures of interest' rates upon which to optimize their portfolios and real investments. Thus financial institutions lack the means to manage cash flow and interest rate risks, and avoid long-term lending in which these risks are paramount. 12 29. Constraints on bank lending. Regulatory constraints on the financial market continue to severely limit the availability of financial resources for private use in general, and for the long-term market rate, mortgage financing in particular. The monetary authorities, on balance, still reserve 25 percent of Bank deposits for public use (through reserve requirements and obligatory purchase of securities), severely constraining the margin for private lending. They also still maintain binding limits on lending rates of financial institutions, inhibiting banks from financing many worthy activities, including housing. 30. Constraints on Long-Term Resources. In the long-term end of the market in particular, CDG is empowered to manage the investment of most of the social security and pension funds (the main generators of long-term resources), as well as of postal savings accounts. In addition, government-owned financial agencies issue obligations guaranteed by the Government, for which CDG is the main investor. CDG manages the supply end of an administered market which controls an annual funds flow of some DH 2,600 million, equivalent to some 16 percent of total financial sector flows; about DH 1,000 of that flow is long-term pension reserves, equivalent to about 50 percent of long-term financial resource flows in Morocco (Annex A-12). CDG's placements include loans to affiliates, government obligations, and obligations of other government agencies. The Government's explicit guarantee of obligations of its agencies, in addition to its implicit guarantee, would curtail CDG's investment in mortgage-backed securities, if such securities were available. Likewise the requirement that insurance companies keep 40 percent of their reserves in the form of government-guaranteed instruments may curtail their investment in mortgage-backed securities. At present, only some DH 500 million (25 percent of total long-term funds, mostly from insurance reserves and some from private pensions) of the annual flow of long-term funds could feed the still nonexistent market for private, including mortgage-related, obligations. 31. While the above constraints remain in place, the Government has taken positive action in the institutional area by establishing a securities commission (Conseil d6ontologique des valeurs mobili6res, CDVM), and preparing legislation for the operation of the securities exchange and mutual funds. But allocating funds to public use through administered schemes and controlling interest rates, more than lack of institutions, constitutes the binding constraint for private business and mortgage lending country-wide. 32. Constraints on the supply of mortgage loans. In addition, as noted in the jointly prepared memoir, a number of policies inhibit banks from offering mortgage loans specifically. Only the Cr6dit Immobilier et H6telier (CIH), a government mortgage lender, has the right to foreclose promptly on borrowers whose payments are in arrears. The other banks must follow the lengthy procedures provided under the Code of Civil Procedure. As for interest rebates (government transfers to compensate the mortgage lender for charging below-market rates for low to mid-price houses), these are not available to all banks, and they may not be as effective in reducing the risk of default as a rebate paid at the time of the purchase and applied toward the borrowers' down payment. 33. Affordability. The above financial constraints, plus the high price of dwellings driven mainly by land costs combine to make it extremely difficult for the population of modest income to finance dwellings considered suitable, on current market terms. As an example, a house of DH 200,000 with a 20 percent down payment and a 13 percent mortgage maturing in 13 20 years would be accessible only to a person of a minimum monthly income of about DH 5,600 (assuming a low 0.30 mortgage payment to income ratio). The Government's traditional desire and the expectation realized by a minority under the present system is that this house be accessible to a family with a monthly income of DH 3,000 (level below which about 50 percent of Moroccan families are). To accommodate new urban families below that level of income, the required Government annual subsidy, hidden or upfront, would be at least DH 100,000 per family and DH 5 billion in the aggregate. While this would derail Morocco's economy, continuing to favor a few with so much under a tight budget is politically untenable in the medium-term. From the mortgage perspective, thus, market finance is a small part of the story; decreasing housing costs, even through decreasing standards, will be necessary to let sustainable, market mortgage finance do its work. 34. Subsidies. Supply- and demand-related policies combined result in substantial total subsidies which on balance appear nontransparent, and regressive. The table below presents a rough estimation (Bank's) of annual public subsidies in the housing sector, which in total amount to DH 3,5 billion, equivalent to about 5 percent of the Government budget. None of those subsidies can be termed fully transparent; only some 25 percent of total subsidies appear to benefit primarily the population of low and low-middle income. On the basis of these rough estimates, Morocco seems to have ample margin to increase direct, transparent subsidies to the poor and still reduce total subsidies. The Government plans to reassess subsidy policies upon conclusion of a Bank-assisted study on financial and fiscal aspects of housing to be undertaken by consultants to the Ministry of Housing; this study will include a more precise estimate of present subsidies and assessment of their incidence. 14 Table 3: Estimated Housing Sector Subsidies in 1992 DH Million % Transfers to public developers, benefitting mostly low-income households.* 383 11 Interest rebates, benefitting mostly low-middle income households.' 190 5 Tax benefit from deductible interest, benefitting mostly high income 156 4 households.a Subsidies embedded in below-market price of lots aimed at low-income 280 8 households.b Subsidies embedded in below-market price of lots aimed at higher income 336 10 households.c Subsidies embedded in below market rents for public housing, benefitting 648 19 several categories of household income.d Tax exemptions and expenditures, benefitting mostly land owners and land 1,501 43 and housing developers.0 TOTAL 3,494 100 a From Annex F b Assuming 8,000 lots of 100 m2 sold at 50 percent of market price of DH 700 m' c Assuming 16,000 lots of 200 m' sold at 85 percent of market price of DH 700 m2 d Assuming 54,000 houses of average value of DH 200,000 rented at half a market monthly rent estimated at 1 percent of house value 15 M. HOUSING STRATEGY A. Context 35. Increasing the private sector's role in land development. Both for the near future and the long-term, under any set of rational housing sector objectives and alternative policies, Morocco must, first and foremost, leave the development of lots mostly to the private sector. The private sector is much better suited than the public sector to give customers what they want and to extract from those with means the total price of satisfying such wants. With the risk of disappearance upon failure to satisfy customers and recover costs, private developers are natural maximizers of value and chargers of full market prices. With existence almost guaranteed, public developers would tend to give customers what seems aesthetically and politically appealing, indulge in below market pricing and, at best, aim at minimizing costs. When the process of urbanization was starting in the 1980s and Government land was plentiful, public land developers could succeed in providing lots to all classes of customers by and large recovering costs, and such success may not have been replicated by an unsophisticated private sector. But the world of the 1990s and beyond is different: public land is more scarce and Morocco's level of urbanization has yet to peak with consequent demands on urban land (para. 7). Under these demanding conditions, there is no room for waste and avoidable giveaways; every dirham spent should produce full value that generally commands full prices in the free markets that attract capable entrepreneurs. To a large extent, if unavoidable, public developers should only provide fully developed lots to the very poor (poorest 15 percent of the population), as such a market would not attract private supply (Bank 1983 Report, Informal Housing: Upgrading and Prevention Policies and Programs already recommended to limit Government-sponsored land development projects to low-cost land development). 36. Strategy objectives. To satisfy needs profiting from market incentives and saving public resources, the overall goal of a housing sector strategy ought be to provide access to decent housing for households as a function of income in competitive markets, and enable the public sector to help the most needy households obtain a house with a minimum outlay of public funds. In accordance with the above housing sector shortfalls and constraints, such strategy should pursue the following objectives: (a) increase the supply of residential land, to be monitored by the number of hectares opened annually for urban use; (b) contain the cost of developed urban land, to be monitored by the ratio of land costs to housing costs; (c) increase the supply of low income houses, to be monitored by (i) the number of low cost houses built annually; and (ii) the ratio of illegal houses to total houses built; (d) increase mortgage financing, to be monitored by the ratio of the outstanding balance of mortgage loans to total loans of the financial system; and 16 (e) albeit more an instrument than an objective per se, but most important, increase the participation of the formal private sector in the supply of lots, to be monitored by the ratio of urban lots annually supplied by that sector to total lots supplied. 37. This report develops a strategy to be gradually put in place during 1995-2000 and to prevail through 2005. Such medium term strategy, however, must be consistent with a vision of the housing sector and the overall economy for the next 15-20 years, and thus avoid erecting barriers to such a vision and instead lead to it as possible. Strategic elements of this long term vision include the following: (a) the private sector will predominate in supplying building lots and houses to middle- and upper-income households in free markets; (b) local governments of urban metropolitan areas and their agencies would be responsible for functional urban planning and for the delivery of serviced land provided with primary, mainly off-site, infrastructure; (c) consistent with the above, most public land developers will largely phase out; (d) the Government will retain responsibility for formulating housing sector policies, provide transparent subsidies to low-income households, and rely on the private sector for the delivery of dwellings whenever possible; and (e) Morocco will significantly decrease the ratio of public expenditures to GDP, and significantly increase the domestic savings rate (becoming less dependent on importing capital and exporting labor) mainly through privatizing the management of pension funds (para. 57). 38. The General order of priorities. To formulate a medium term housing strategy, moreover, we must take into account two guiding principles. The first is what the general order of priorities is. Among priority actions, those aimed at eliminating constraints on the supply of urban building lots rank higher than those designed to stimulate demand. Raising subsidies and facilitating loans without removing supply constraints would fail to increase the availability of lots and houses, and, instead, further increase their prices. Substantially increasing housing finance without first reducing supply constraints, moreover, may cause significant mortgage loan losses in the future. Financing would help further boost property prices, but such inflated prices eventually will have to come down, reducing the guarantee on mortgages. Bank soundness, thus, calls for addressing land and housing supply constraints, before ambitiously pursuing to boost real estate finance. Still, with supply constraints removed, availability of financing must immediately follow, as lack of such finance would suffocate the supply response. 39. The medium-term role of public developers. The second guiding principle is that medium term strategies should be realistic in terms of the path of sector development, and, to the extent possible, profit from positive aspects of the present phase of that path. This in particular calls for existing public developers' being responsible for developing land on the urban fringes mainly with on-site main infrastructure to turn to the private sector for lot development, and fully developing lots only for the very poor (primarily the 15 percent of the urban population with lowest income). As further explained in paras. 47 and 48, public developers could do this 17 on a full cost recovery basis under several types of operations that result in some recovery of land from original owners to pay urban expansion costs and house the poor; such operations would even maintain cross-subsidization of lots for the poor, but out of the benefits obtained in wholesale tract operations, not retail lot operations with higher-income households. The public sector must take the lead in expanding main (on-site and off-site) infrastructure, as benefits from such activity would generally overflow the area of interest of private developers. Over the next few years, municipalities and R6gies, more likely than not, will lack the resources to undertake this role as well as the political objective and drive to fully recover costs from the users of services. The Government, for its part, ought not pay for main infrastructure out of budget funds, as such payment would drain national resources to the benefit of present land owners. Thus public developers afford the best present alternative to assuring expansion of urban land serviced with main (primarily on-site) infrastructure, increase the participation of the private sector in land development, and avoid deriving further subsidies from the central Government to house the poor. 40. The following paragraphs, in the sequence of the process of supply, describe the regulatory, operational and supporting measures necessary to improve the supply of dwellings, and also provide their rationale. Such policies are to be put in place during 1995-2000, and remain in place at least until 2005. Providing a time frame, table 4 summarizes all strategies recommended with certainty for this period, including the more notable of complementary financial sector policies, and changes in the housing sector institutional set up. Table 4: Summary of Proposed Measures Components Actions to take during Actions to take during 1995-1997 1998-2000 1. Streamline and operationalize - Accelerate review and approval - Introduce land development urban planning -of urban documents. phases in urban planning - Introduce lower standards for documents. lots aimed at low income - Optimize operational urban households.' standards and revise regulations - Establish system for public to give more flexibility to the sector to launch ZAC and private sector.' replotting operations, and to - Combine land survey with take private land to recover development plan, to orient costs of, and benefit from, private sector decisions. urban expansion. a 2. Develop information tools. - Start producing charts of market prices. 3. Let developers design - R6gies permit developers to infrastructure works. design and carry out works, subject to Regie standards. 18 Components Actions to take during Actions to take during 1995-1997 1998-2000 4. Modify role of public land - Steer them towards providing - Launch ZAC and replotting developers. primary infrastructure and fully operations in the main cities.' developed lots only to the very poor.s b - Launch initial sales of tracts with primary infrastructure to private sector developers in the main cities." - Promulgate requisite laws and plan for ZAC and replotting operations.' 5. Improve housing sector - Lower import duties for key - Simplify real estate taxes, taxation. construction materials.b expand the territorial base, and update such base in tandem with inflation. - Eliminate real estate tax exemptions and related nontransparent subsidies. 6. Deregulate rents. - Pass law allowing the rapid - Enact law streamlining rent eviction of tenants in arrears.' control. - Sell public rental houses. 7. Liberalize interest rates. - Suppress limit on lending rates. 8. Facilitate mortgage finance. - Issue CDVM's regulations on - Lower somewhat further the the issue of Mortgage bonds. percentage of reserves insurers - Withdraw government explicit must place in government guarantee on obligations of guaranteed obligations. public agencies.b - Extend rapid procedures to foreclose mortgages to all financial institutions." 9. Improve the institutional set - Transfer urban management up. responsibilities from the communes to the communaut6s. - Withdraw land development mission from urban agencies. These measures require further study before, and some gradualism in, implementation. b Measures which best combine facility, convenience, and urgency, and thus require immediate action (early phase of 1995-1997 period). C Action which is the backbone of the proposed urban land management strategy. d Financial sector liberalization measure which is essential for mortgage markets to work. Other measures to liberalize money and capital markets should facilitate mortgage finance. 19 B. Policy Recommendations 41. Streamlining urban planning. The importance of urban planning in the land delivery system warrants the adoption of a number of actions to some extent called for the urban law of 1992, which is not yet fully operational. Main actions are the following: (a) Streamlining the preparation, approval, and implementation procedures of urban development documents, to reduce uncertainties of land owners and potential developers, through: (i) decentralizing to the Governors the approval of urban plans for small cities, avoiding the lengthy process of consultation among central agencies; and (ii) sending land use plans once only to all parties concerned (ministries, municipality, citywide government), consider it cleared with lack of comment within two months, and modify it once only before final approval. (b) Preparing a timetable for implementation which differentiates areas to be developed in the short term from those to be developed in the future, to better orient the decisions of land owners and developers (under implementation now). (c) Combining the land survey with preparation of the development plan, to focus the attention of public sector agencies on opening more appropriate areas for development through more suitable instruments for facilitating the private provision of lots (the authorities will implement this for future urban development plans). 42. Optimizing urban development standards. The effective use of land resources is largely dependent on urban development standards. The optimization of urban standards along the following lines is recommended: (a) Revise regulations and standards to give more flexibility to developers, particularly adopting a floor/area ratio of about 2.5, and reducing minimum width of streets to about 8 meters, except for main roads; (b) Use land more efficiently in public developments designed to provide lots for low- cost housing, by increasing density to the level allowed by land use plans and the land utilization ratio to about 0.65; and (c) Plan for differentiated development (lower initial standards for lots aimed at low- income households) to make property ownership more affordable by incurring infrastructure costs over time. Still, the optimization of these standards calls for further analysis which the Moroccan authorities will carry out with the assistance of consultants under the Bank-assisted Land Development Project. The schedule of this study would permit optimizing the standards after 1997. 43. Operationalizing urban development instruments. Under prevailing conditions, it is necessary to design and to apply the operational urban development instruments described below, which respond to different situations of land tenure. 20 (a) The establishment of comprehensive housing development areas (ZAC) for major projects where the number of land owners is limited, with negotiations between the private developers and the public sector to define not only the physical features of the project but also the allocation of responsibility for financing infrastructure between the two partners. (b) Land replotting, or mandatory urbanization of private land (a far better alternative to expropriation for the owners), which is more appropriate than voluntary consolidation in situations when property is too fragmented in the target area or the development of large tracts of land is called for. In both cases, a mandatory system should be instituted for the free transfer of land to the local governments or public developers (perhaps in the order of 40 percent) for the construction of infrastructure and other public facilities. Moreover, the local government (or public developer) could recover the cost of primary infrastructure and make a profit by selling tracts, and then use this profit to provide lots for low-income households in the retained tracts (paras. 47 and 48 explain how to operationalize these instruments and what results can be achieved, respectively). 44. Management of the land market: instruments for intervention. The Government, as a safety valve against major price escalation upon urban expansion plans, ought to establish a system for exercising preemptive rights (right of public entity to purchase at the same price offered by a private agent) in areas to be developed or renovated in the near future. In this respect, whenever the Government creates an area that is subject to rights of preemption, it should concurrently designate a reference price based on market value as of the date the area is established, but not including prospective rises in connection with the announced land development project and the attendant off-site services to be installed. Still, as discussed in detail in Annex D, preemptive rights is a sensitive tool, which should thus be used sparingly. 45. Developing information tools. To address the relative lack of transparency in land markets and support the above preemption tool as needed, in the initial stage, the Government and the private sector should jointly develop a simple system to monitor land values based on appraisals by professionals in the sector (government agencies, real estate agents, private and public real estate developers), and to produce charts of the reported land prices. These charts should be widely distributed and used to check prices applied in expropriations, and particularly to check the prices of transactions, for sales both of unregistered and registered tracts. In the second stage, publication of all real estate transfers and land transactions could be envisaged. The main goal of this measure would be to enhance the transparency of markets. 46. Expansion of infrastructure. The provision of infrastructure is an essential component of the land delivery system. It requires the following changes: (a) R6gies should let developers, including those working in the private sector, design infrastructure themselves, with R6gie standards and under its supervision (para. 16), a measure which would expedite the implementation of urban developments; and 21 (b) The Government should establish a system for the public sector to recover the cost and benefits of urban expansion in the form of land taken from private owners deriving benefits from such expansion (para. 43). 47. Changing the role of public land developers. In the context described in paras. 18-21, it is essential to redefine the mission of the public developers (SNEC, ANHI, ERAC) by, among others, eliminating the present cross-subsidization as a system for producing lots for both low-income and well-to-do families as fast as prudent implementation allows (para. 49). Under the framework discussed above (para. 43), public developers, acting as wholesale land entrepreneurs, should increase the supply of urban land by equipping it with main, primarily on- site infrastructure, including main roads and major water, sewerage and electricity networks. Still, as they do today to a large extent, public developers over the medium term may often provide even off site infrastructure and recover related costs through their operations. Private developers would generally put in place minor roads and secondary and tertiary service networks. Municipalities and their agencies, who over the long term should place off site infrastructure fully recovering related costs from the primary beneficiaries of urban growth, may take some time to muster the resources and control the mechanisms for fulfilling such responsibility (para. 37 and 79-d). Public developers, moreover, should also provide developed lots to the poor, as no substantial alternative is foreseen in the medium term. These recommendations are consistent with a 15-20 year vision in which public developers should largely phase out, and the private sector should dominate urban land development; public developers dedicated primarily to install main infrastructure would become lean and thus easier to phase out when indicated. The public developers, as noted in the jointly prepared memoir, should fulfill their mission in the following ways: (a) on public or private land which is easily obtained, install primary infrastructure and sell large tracts at a profit to private developers through competitive procedures, keeping a part of the land to develop lots for the poor; (b) in areas divided among a few owners, provide primary infrastructure under an agreement with private owners (ZAC operation), which includes acquisition of part of the land for the public developer to recover costs, make a profit, and develop lots for the poor; and (c) in large areas divided among many owners, launch land replotting operations, under which, by administrative decision, the public developer provides the main infrastructure, gives their tracts to the owners as restructured and reduced, and also takes part of the land to recover costs, make a profit, and develop lots for the poor. 48. The above operations have well established precedents, and good prospects of being financially viable. The sale of tracts to private developers is just a modification of what public developers now do in Morocco; it should be financially viable and maintain cross- subsidization to the extent that the sale of tracts to private developers through competitive procedures at what the market would bear would compensate public developers for the loss of the tract-to-lot conversion margin. The ZAC operation has proved effective, mainly in France. The replotting operation has been used successfully mainly by municipalities in many countries of the world, including Germany and South Korea, whose systems inspire the present proposal. In Germany, land retained by the municipalities for public uses (roads, public buildings) and cost 22 recovery can go up to 30 percent of total, and such municipalities aim at no more than recovering costs. In South Korea, land retention has gone to 40 percent, and municipalities have benefitted mainly through the sale of commercial plots. In the Moroccan context, a benefit to apply to social housing appears possible and warranted. Annex A-13 presents an hypothetical replotting model, but one built on the physical parameters and estimated market prices corresponding to the Sidi Abdellah SNEC's lots sale project. It is assumed that owners recover about 60 percent of the raw land area in land with primary infrastructure, and that the public developer allocates about 15 percent of the total area to main roads and about 25 percent to recover costs, shelter public services, and develop lots for the poor. On the basis of estimated current prices, land owners and potential private developers retain substantial profit margins, and public developers meet all costs and furnish around 2,000 lots of 80 m2 for the original 350 hectares of the total project; the small plots should concentrate benefit in the lower half of the poorest third of the population, and foster construction of houses of some 60 m2 rather than the predominant 70 m2. Somewhat increasing land taxed away from owners may significantly increase the number of lots for the poor, while requiring to use part of large private tracts for lots of 100-120 m' may result in low-middle income households purchasing these lots. The present example shows a major disengagement of public developers whose development costs could come down to about half of those under the present system. 49. Changing the role of public developers as proposed calls for the fulfillment of three conditions. The first is that those developers and government authorities proceed prudently with, but seek firmly, such a change. On the one hand, public developers to a large extent should maintain planned traditional operations to avoid disruptions of supply, and approach commitment to the new operations on the basis of testing and learning. On the other hand, the. government would have to monitor the change as a national enterprise, and public developers expeditiously develop substantial operations of the new type. Carrying out a few major operations in main cities where land is available and urban expansion necessary would be a good way to start the reform (para. 81). The second condition for the success of the reform is that, within the framework of rational planning and on a case by case basis, the authorities in charge of urban planning allow for land development through the above operational tools. The third condition is that public developers have access to some additional financing through the financial system, as financing requirements would increase with increased supply of urban land and decreased advances from higher income families that would not be forthcoming under the new operations. 50. Emproving land taxation. As a longer term solution aiming to simplify real estate taxes and foster land sales, Morocco should consider replacing the various real estate taxes based mostly on rental values with a single tax on the market value of land; this option requires careful assessment, which could be undertaken in a study in progress sponsored by the Ministry of Housing (further discussion in Annex D; relevant literature in Annex F). Still, for the immediate future, Morocco has margin for increasing tax revenues and fostering sales with minor adjustments to the present system on three counts. First, as long as rural land is exempt (exemption which needs to be reconsidered), the urban perimeters for tax purposes should include both the land use plan area regardless of land use and urbanized areas out of such a plan, as zoning and urban use both boost land prices. Second, the Central Government should annually update the base of real estate taxes in tandem with inflation (the authorities should continue reassessing property values every five years). Third, taxes on nonconstructed urban 23 land should be reviewed and revised as necessary to be consistent with the preceding changes and, as a percentage of implicit property values, have rates no lower than those on constructed land. 51. Other measures to improve housing taxation. These, perhaps within the context of more comprehensive reforms, should include: (a) Lowering import duties as much as permitted by the need for customs revenues and requirements of Morocco's trade agreements (cement and reinforcing rods appear to be good candidates, since tariffs primarily protect local industry and thus lower tariffs would not lower customs revenues); and (b) To the extent possible, eliminate the deduction of taxable income for mortgage interest payments and tax exemptions on real estate profits, and use the resulting increase in fiscal revenues to transparently subsidize the poor. 52. Deregulating the rental market. In the rental market, to encourage investment in rental housing and thus release pressure on the public sector to help most low-income families own a house, the following measures are recommended: (a) enacting the law establishing rapid eviction of tenants who are late in paying their rents; (b) enacting a law authorizing annual rent increases on the basis of the consumer price index, and abrogating the requirement to seek court approval for increasing rents; and (c) selling public rental units by taking advantage of rent increases and therefore of higher property values resulting from enactment of the above laws. 53. Need for financing. Although housing supply constraints dominate those on demand, increasing supply calls for not only first removing supply constraints, but also for increasing financing. Loans to land developers, housing promoters, and mortgagees must lubricate the supply system and foster sales prospects, for the housing sector to deliver a substantially higher number of dwellings. 54. Liberalizing financial markets. More liberalized financial markets should largely precede the sound development of mortgage finance and determine the stages of this development. Among the many actions required to liberalize financial markets (dealt with through dialogue between government authorities and the Bank team working in financial sector reform -- para. 28), one essential strategic action that would specially facilitate mortgage financing ought to be emphasized: financial institutions should be free to set lending rates (measure considered under the Financial Markets Development Loan). Without such freedom, mortgage originators would screen out many risks otherwise worthy at rates higher than the imposed limit. Other measures pertaining mainly to the money market would aim to further liberalize interest rates, reform regulations of financial markets, reduce the Government preferential access to credit, accelerate the use of indirect monetary policy instruments, and enhance integration with world financial markets. Still, as further explained below, a substantial 24 long-term debt market with an also substantial mortgage segment would most probably require that the pension system--a natural generator of long-term resources--be fully funded and privately managed, a proposition that is not foreseen for the next few years (para. 57). Without this reform, mortgage funding through the internal market will be modest. Thus, for the next few years Morocco's approach should consist mainly of establishing a more level and secure playing field for the participants in the market for long-term resources; it may also foster some directed credit if needed to allow housing supply to respond to market stimuli, but avoiding the institutionalization of directed credit which may later undermine the development of free capital markets. The following paragraphs describe policies consistent with this approach. 55. Facilitating mortgage finance. As noted in the jointly prepared memoir, the Government should take the following measures to enable potential mortgage lenders (banks with mainly short-term deposits) to obtain long-term funds from the institutions which naturally generate such funds: (a) Develop a market for mortgage securities, in a manner similar to that proposed by the commission which studied the reform under the auspices of the Ministry of Finance and Banque Al-Maghrib. According to this proposal, the banks would issue securities that would be guaranteed by a portfolio of mortgages. The banks, who would service and maintain the mortgages in their balance sheets, would directly place the issued bonds with the originators of long-term resources (insurance companies, CDG, private pension funds, mutual funds). Those bonds, however, ought not to carry the guarantee of the Government (which the above commission has recommended -- following paragraph). CDVM's regulations on the requirements to issue and trade this paper (1.2 mortgage to bond ratio, 0.80 loan to house value ratio) should be issued shortly. (b) Withdraw the Government's explicit guarantee on securities issued by government agencies, generally purchased by CDG (measure considered under the Financial Markets Development Loan). This measure would somewhat level the playing field among issuers of debt, enhance CDG's sensitivity to interest rate differentials and, as a result, give mortgage- originating banks some access to CDG's resources. (c) At least when a rating agency is in place (so that investors could purchase bonds issued not only by supervised banks, but from other corporations as well), further liberalize the investment guidelines for insurance companies. 56. The above measures imply major choices among alternatives, and along time. The establishment of the bond market in particular includes three choices which call for justification. First, the recommended solution does not include a government guarantee that the bond-holder will be repaid by the mortgage originators. Such a guarantee would distort the pricing of risk and the Government may ultimately suffer the consequences of investors' lack of interest in taking bonds in accordance with the soundness of mortgage-originating banks. Second, the solution does not involve securitization, which requires transfer of the mortgages to a newly created intermediary (banks would have mortgages off their balance sheet) which would issue either securities representative of its mortgage pool or plain mortgage backed securities to be placed among providers of long-term funds (insurance companies, CDG, or pension funds). The transfer of mortgages and the attendant elimination of all risks to the 25 originating banks would either place a heavy risk burden on the agency intermediating between the banks and the providers of long-term funds, or entail substantial administrative costs to manage that risk. Securitization, like in the United States' Fannie Mae, requires a major infrastructure (appraisals, title security, legal responsibilities for misrepresentation) that may take decades to put in place, without clear justification. Third, although possible if banks were to issue mortgage bonds, such a recommendation (at least for the immediate future) does not include the establishment of an intermediary who would take the bank bonds and place its own among the generators of long-term funds. Without majority government ownership (such majority ownership would amount to an implicit Government guarantee and ought to be avoided), that intermediary would have limited usefulness. Its main positive, but modest contribution would be to transform small bank bond issues into more liquid major issues; but the possible premium on such issues may fall short of the capital and management costs of having the intermediary. Still, when housing supply constraints are substantially overcome and finance constraints appear binding in spite of liberalizing measures, the issue of whether the establishment of the intermediary is warranted ought to be revisited. 57. The other major short-term choice is limiting the liberalization of long-term funds mainly to stop giving government guarantees on bond issues by public agencies, instead of privatizing pensions. As mentioned above, over the longer term, say around 2005, the Government ought to probably leave the management of pensions for public and private sector employees to private pension funds. Private pension programs, funded of necessity, are the best feeders of capital markets and, substantial research suggests, major boosters of national savings rates (Annex F). This, however, would require substantial study, and call for major political choices. Furthermore, there is some base for first significantly liberalizing land markets before fully liberalizing long-term, capital markets. Land prices supported by supply constrains and fueled by financing could raise so much above sustainable levels that their inevitable coming down may result upon substantial losses to investors and financiers. 58. Extending rapid foreclosure rights. Furthermore, to encourage banks to increase their provision of mortgage loans, the right to rapidly foreclose on mortgaged property, which is reserved exclusively to the CIH at this time, should be extended to these lenders. 59. The above reforms, including further liberalization of the overall financial sector, raise the question of the extent to which banks will start financing mortgage loans to middle- income families with long-term funds from insurance companies and other generators of long- term funds. Such an extent, which is difficult to determine reliably, will vary in accordance with the extent to which the Government pursues the above reforms. Only with the privatization of pension management, not envisaged in the medium-term, capital and mortgage markets would probably boom. Still, establishing a more level playing field among possible mortgage originators and among users of long-term resources -- plus further liberalizing financial markets, including lending rates - should increase mortgage lending noticeably; some insurance and pension reserves could flow to banks, and.these would be more prone to finance developers and middle-income households. A significant limiting factor would still be that, with the above mortgage bonds, banks would maintain mortgages in their portfolios, running the cash-flow risk of imperfect maturity matches, and thus would choose to limit their mortgage exposures. Bringing down the banks to the low-middle income segment of the market, for its part, would require some subsidy to reduce lending risk (forthcoming para. 60). Yet on balance, with 26 housing supply constraints dominating those on demand, a modest increase in mortgage lending should suffice to enhance market performance to its limit for the next few years. At the beginning of land market liberalization, some directed housing credit, probably from foreign lines of credit, may also have to be available to prevent that lack of financing makes supply abort. For later, the Government will have to decide on measures such as privatizing pension management and, maybe, helping establish a mostly privately owned mortgage intermediary to struggle for mortgage funding if the placement of mortgage bonds proves weak and both banks and generators of long-term funds prove intent on capitalizing such an intermediary. 60. Possible financial facility for low-middle-income families. For the longer term, a measure worth considering is the establishment of up front Government grants to be applied to the initial down payment on a home purchase. This system, which has been successful in Colombia and Chile, lowers the minimum income needed to qualify for a mortgage loan, thereby reducing government intervention in the provision of public housing. This measure, however, should be examined more comprehensively and, if applicable, should not be implemented until the supply system has adjusted to the liberalization of land and housing markets. 61. Targeting additional assistance to low-income families. While the operational urban development instruments designed to increase the supply of land will help to provide building lots and, potentially, housing for the poorest households, the financial reforms will be of no direct use to these groups. This is inevitable, since any attempt to require that financial institutions and private entrepreneurs assume unwanted risks will fail and will cause other aspects of these reforms to fail as well. Hence, for when the Government has the necessary resources, two additional measures, requiring in-depth examination to carry forward, are suggested: (a) implementation of a system of direct assistance to very low income households who would be identified by local governments as eligible for rent subsidies in the form of vouchers redeemable by the landlord and financed jointly by the local government and the Treasury; and (b) establishment of an assistance program for low-income households who are building or improving houses in slum areas that are being upgraded. 62. Immediate Measures. At present, the specific measures which appear to best combine facility, convenience and urgency, and thus require immediate action (early phase of the 1995-1997 period), are the following: (a) establish a system for the public sector (public land developers initially) to carry out ZAC and land replotting operations, and to recover the costs of, and derive benefits from, urban expansion, by taking private land when primary infrastructure is put in place (para. 43 -- legislation would be required, see para. 79-b); (b) redefinine progressively mission of public developers as expanding the supply of land with urban main infrastructure and providing fully developed lots only to the very poor (para. 47 -- may eventually require change in charters); 27 (c) launch sales of tracts with main infrastructure to private developers in the main cities (para. 47-a); (d) lower import duties of key construction materials (para. 51-a); (e) pass law allowing the rapid eviction of tenants in arrears (para. 52-a); (f) suppress the explicit guarantee of the Government for obligations issued by public agencies (mainly taken by the government-owned Caisse de D6p6ts et de Gestion), so as to have a level playing field for other obligations including those collateralized with mortgages (para. 55-b); and (g) extend the privilege of rapid foreclosure procedures on mortgages--now held exclusively by the government-owned Cr6dit Immobilier et H6telier--to all financial institutions (para. 58). (h) main complementary measures of financial sector liberalization, including the removal of limits on lending rates (para. 54). 63. The above set of policies, among others, intentionally lacks one substantial element: a proposal to substantially increase the public sector supply of completed houses for the low income population. This calls for explaining why this is not proposed, and providing a perspective on how the public sector will foster low-income housing under the above policies. C. Low-Cost Housing Prospects Under the Above Policies 64. The strategy of providing mainly lots to low-income families is socially viable, and cost effective. The public supply of completed houses for the low-income population is very expensive and, when done on a substantial scale, generally results in drainage of budget funds through long-term finance and outright loan losses, as the low-income population has no access to market funding. Only predominately urbanized countries with low population growth and intermediate income levels, not the case in Morocco, can indulge in this expensive proposition. Morocco's public sector has rightly shied away from the provision of full houses to the low- income families, concentrating instead in the provision of lots, and should not massively undertake such provision now. Providing full houses to a minority of the low-income population, for its part, appears unequitable. 65. The above supply and demand strategies, moreover, should effectively foster the provision of low-cost housing to the low-income population, with economy of resources. Liberalizing the rental market, to begin with, would enable investors to construct low-cost housing for rent. Adjusting urban standards and other policies for decreasing housing costs would go a long way towards making legal housing more affordable to the low-income population, before resorting to subsidized public sector supply or other government grants. Public sector supply of small lots in a number equivalent to 15 percent of new households formed annually would facilitate the supply of housing to about double that share of low-income households, as lots accommodate comfortably two houses and sharing of construction costs and informal financing are well established. For later, the above mentioned direct grants towards rent 28 payments and house improvement would further enhance the housing conditions of the urban population. With these measures, the Government, indirectly through facilitating policies and directly through granting subsidies, would help house a number of low-income households at least equivalent to one third of new households formed per year. The nature of the policies combined with some targeting of the programs could, with no major problem, ensure that the benefitted households are those of lowest income. With these conditions, fostering the public provision of full houses seems unwarranted. 66. Under the proposed policies, targeting low-income families with subsidized lots or other possible programs should not pose a major problem. At present, the Ministry of Housing identified the list of families living in selected shantytowns who will obtain subsidized lots near their present dwellings, and furnishes such lists to the public developers. Public developers should easily extend their choice to low-income families sharing a house in more stable, but illegal settlements. What has blurred the targeting of low-income families is that most public developers provide subsidized lots to low-middle-income families. Under the redefined mission of public developers, they would fully develop lots of up to some 80 M2, which should help maintain their lot-supply focus on the low-income families only. Low-middle- income families would get their lots from private suppliers motivated by appropriate urban standards, zoning and prices (para. 48). 67. To provide some 15,000 lots to low-income families annually after year 2000, the total annual subsidy would be about today's DH 1.1 billion (assuming market prices of DH 1,500 and subsidized prices of DH 600 per m2 -- Annex A-13). Such a subsidy would mostly come from public developers' benefits from installing main infrastructure in tracts for the private sector, at no cost to the Government budget. The amount of that subsidy would be less than one third of total (overt and hidden) present estimated subsidies in the housing sector. D. Benefits from Recommended Policies 68. The above policy package for the medium term should pay handsomely in terms of national welfare. At the micro level, the benefit will mainly appear as a decrease in the cost of dwellings to the final user at no resource cost to the economy or the Government. As documented in Annex A-14, the cost of a today's DH 200,000 house could come down about 25 percent. Main savings would come from reduced land development costs and prices of construction materials. To the extent that the policies are effective, current land owners would bear a loss of wealth and producers of construction materials would bear a reduction in the stream of revenues; but this is as it should be, as land values are inflated by distorting policies and enterprise revenues inflated by distorting tariffs. 69. Reduction in costs, plus adjustments in housing quality forced by letting markets work, would make houses much more accessible to households with modest incomes. As an example, a house reduced in price from DH 200,000 to 150,000 would become accessible to a family of DH 4,500, compared to a family of DH 5,600 before, without need for subsidies. With a fully liberalized financial sector, financial institutions would find the long-term resources and the inducement to finance mortgages for income segments now deemed unacceptable risks. 29 70. Seeing the sector perspective, the main benefit, albeit difficult to quantify reliably, may prove fiscal. The operational instruments of land development could recover substantial land for society, re-appropriating the social externality caused by urban expansion that otherwise would have benefited current owners only. Other benefits also difficult to quantify reliably would be those obtained by increased efficiency of the development of lots and housing construction by private developers. Land liberalization in particular may permit larger housing projects completed at decreasing unit costs. 71. From the macro perspective, most of the potential benefits appear substantial. For the total economy, welfare, not necessarily gross domestic product at market prices, would increase, as most benefits from reforms come from increased efficiency and undoing regressive distributions of wealth. The financial policies in particular, moreover, would contribute to increasing the level of financial intermediation and making financial markets more efficient, which would foster Morocco's domestic savings rate, productivity, and economic growth. On a more concrete, but limited ground, the public sector could recover in the order of DH 2 billion by eliminating unwarranted tax incentives (Annex E). 72. Although the proposed reform is far reaching, the above substantial benefits would almost come at no cost in social, economic, and budgetary terms. For as long as land remains somewhat scarce for the population of lower income, families obtaining a subsidized lot would be able to swap the right to build a second floor for some cash, perhaps with a relative, so the Government would have little pressure to provide houses, an expensive proposition, on top of lots. The population of middle and high income would not feel the absence of subsidized lots, because land prices in general should come down significantly. With controlled urban. expansion by public developers and relaxation of urban standards, the total cost of urban expansion at least would not increase. Streamlined public developers would be self-financed and thus avoid burdening the central budget. The Government may use fiscal resources from additional tax revenues and savings from the proposed elimination of tax exemptions to give up front grants to the poor, but entirely at its will, under reduced social pressures to perform. 73. At the horizon of 6 years, when the reform could realistically be largely in place given the political. will to do so, its positive effects can be put into perspective by comparing how many more acceptable dwellings could be made available without increasing the level of economic and fiscal resources applied to housing; the order of magnitude is some additional 35,000 (from a base of 100,000), a combined result of cost savings and recovery of unwarranted tax exemptions. This increase would almost suffice to cover incremental needs coming from population growth; a growing economy with its generation of additional resources, private and public, would suffice to even start replacing inadequate with adequate dwellings on a significant scale. 74. The table and graph below compare housing sector performance and sources of supply of urban lots in 1985 and 1995, with those estimated for year 2,000 (when most new policies would be in place). On the basis of realistic assumptions that underlie the table, Morocco should be able to substantially improve housing conditions with a modest increase in housing investments that even minimal economic growth would accommodate. Morocco's envisaged more dynamic growth should further enhance potential housing sector performance. Privatization of the supply of urban land, for its part, should be substantial; for year 2000, formal private sector developers would more than double, and public developers about halve, 30 their expenditures in land development (estimated on the basis of the present table and Annex A-13). Table 5: Housing Sector Prospects (Values in constant DH Million of 1992) 19851 % 199-b) % 2,000) % Number of Houses built 80,000 100 115,000 100 126,000 d) 100 Number of legal houses built 49,800 62 71,300 62 94,500 75 Number of low-cost houses built 21,200 27 31,000 27 41,600 33 Number of low-cost legal houses 13,500 17 20,000 17 32,700 26 built Number of lots supplied 42,300 100 51,500 100 63,000 100 Number of lots supplied by formal 7,300 17 8,500 17 22,000 35 private sector Number of public lots supplied to 9,000 21 12,000 21 13,000 21 the poor Number of lots supplied by 7,000 17 10,000 17 9,000 14 informal market Housing Investments 20,000'* 3,0000) 100 25,200') 100 Explicit and implicit housing 4,200 e 18 2,000 h 8 subsidies Financial Institutions Private Sector - 114,000 ' 100 153,000 0 100 Loan Balance Mortgage loans outstanding 8,000 7 20,000 13 a) Data from Attachment A-1, applied to the mid point year. b) Estimated on the basis of 1985, assuming a growth of 4% per year and similar structure of supply. c) Expected structure of supply d) To cover population growth of 4.3 p.a., assuming a family size of 5.5. e) Assuming an average cost of DH 250,000. f) Assuming an average cost of DH 200,000. g) Escalating data on Table 4 at 6% p.a. h) Allowance which easily covers transparent subsidies. i) Escalating data of 1992 at 6% p.a. 31 Lot supply in Morocco for the next 15 years 100000 Bidonvilles 80000 . informal market 0 60000 - formal private supply E4000 0 Public supply for high and 20000 middle income 0 U Public supply for low- 1985 1995 2000 2005 2010 income E. Institutional Recommendations 75. To ensure the success of the above policies, the institutional arrangements must facilitate the work of public officials and embed incentives to make them pursue such policies. Three possible changes, in addition to redefining the mission of public developers mentioned above, are discussed below. 76. First, the responsibility for managing urban growth regulation at the local level (within the tutelage of the Central Government which will be necessary for many years) should be vested in city-wide governments or communaut6s, not in the micro municipalities or communes into which major cities have been divided after the municipal fragmentation of 1994. It is apparent that developing and approving urban plans at the level of neighborhoods could be highly inefficient. The present arrangement, moreover, deters effective decentralization envisaged for the long term, as micro communes have less possibility of becoming self reliant in urban affairs than city-wide governments. City-wide responsibility for urban management is the natural step toward facilitating enhanced local independence from the Central Government in the future. 77. The second change pertains to the Urban Agencies. These agencies should preferably lose their land development power, as it conflicts with their more useful and proper role of supporting urban regulation, which they should preserve. Considering that stronger, city-wide governments ought to become responsible for more autonomous urban management, moreover, the creation of new Urban Agencies beyond Casablanca and Rabat, the capital city, seems unwarranted. As a second best, the Urban Agencies may continue with the current practice of only carrying out pilot developments as a means to support urban development plans, a limitation which would require careful monitoring from the part of the Government. 32 78. Third, for the longer term and within the context of state design, the placement of the entity in charge of urban management in the structure of the Government ought to be carefully considered. Ideally, to foster accountability, urban management should be vested in a department or agency dedicated exclusively to such a function. Thus, the Ministry of the Interior, with its several important, demanding functions, may not be the best place; nor is the Ministry of Housing as the mission of housing the poor may conflict the more general one of regulation. A High Commissariat under the Prime Minister appears a promising alternative. F. Action Plan to Carry Out the Reform 79. While some of the above policies could be implemented shortly with minimal effort of government officials, some others require major tasks, including preparation of legislation and studies geared to implement those reforms. Major tasks foreseen now (mostly considered in terms of reference of studies sponsored by the Ministry of Housing with Bank assistance) include the following: (a) Establishment of new planning procedures and standards. This will require the preparation and approval of a law changing the process of approval of urban documents by the Government. It will also require a fine tuning of urban standards, and a standardization of urban documents (paras. 41 and 42). (b) Legislation on Land Transfers to the Public Sector and on Land Replotting. Public sector recovery of the benefits of urban land expansion and the operation of land replotting will require a major piece of legislation. Even before the legislation, the Government should determine what fraction of land recovery for social use is appropriate, and what such a fraction means in terms of public establishments, roads, land available for the poor, and land available for public profit in terms of cash from sales (para. 43). (c) Program to Steer Public Land Developers Towards the Provision of Main Infrastructure. The charter of public developers should be revised to clearly establish that their mission is expanding main infrastructure and providing lots only for the very poor (para. 47). (d) Land Supply Projects. A major effort should be conducted to transform present traditional lot development projects in the preliminary stage into land supply projects aimed at putting large tracts in the hands of the private sector and urban lots in the hands of the poor. New projects, moreover, should be prepared in an accelerated manner. A main concern in these designs will be to make sure that at the level of each city the projects will be sizeable enough to impact the market, and organized in a centralized manner, so that confusion among several public developers is avoided (para. 47). (e) Taxation Changes. The establishment of a tax on property values, and the phasing out of the real estate tax code will require substantial work. For establishing that property tax in particular, it will be necessary for the fiscal authorities to incorporate the limits of the land use plans in their systems and establish property values to assess the tax (para. 50). 33 (f) Revised Project of Law on the Establishment of Rents. The authorities should prepare a new law permitting annual rent increases on the basis of the consumer price index, and abrogating the requirement to seek court approval for adjusting rents (para. 52). (g) Feasibility Study of System of Up Front Grants Towards the Down Payment on a House. The possible system of up front grants (to replace interest subsidies) is being analyzed under one of the Bank-financed studies. Main components of the study should include the assessment of the target population with its necessary grant levels, the projections of fiscal cash flows from the phasing out and incoming system, and the design of the organizational arrangements for managing the new grant system (para. 60). 80. To avoid reexamining policies that have already been defined, studies about to be undertaken under present Bank-assisted projects should take into account the findings of this report. The appropriate Government officials and the task teams of the concerned Bank-assisted projects are addressing this matter. IV. BANK ASSISTANCE 81. Beyond the above-mentioned assistance in implementation planning, future Bank operations ought to underpin the proposed strategy. The most immediate follow-on operation should primarily pursue placing land with main urban infrastructure at the disposal of the private sector in selected major cities. Thus, one of the main project components would be funding to put in place main infrastructure in large sections of the cities and also to complete lots for the poor. Such public developer operations, to make a substantial difference in land supply, preclude substantial traditional retail operations for all classes of families in the same cities; substantial traditional operations would otherwise maintain non market competition over private developers. Furthermore, on the basis of the project substantially increasing land supply and underpinning removal of above constraints to mortgage financing, the Bank loan should probably provide some funding for housing directed credit, to ensure financing of the intended supply. Clear and controlled directed credit given to intermediaries or channeled through a fiscal agent is a valid way to cover for transitory deficiencies in the financial system; because of its open honesty, it does not undermine the further development of such a system. 82. A next follow-on Bank operation should mainly fund main infrastructure and lots for the poor primarily through replotting operations of public developers, and maybe a program of up-front subsidies towards the downpayment on low-cost houses financed by financial institutions. Up front subsidies help the private sector housing supply and financing come down to lower-income segments, and reduce Government involvement in the supply of housing. Furthermore, such a follow-on Bank operation, if private sector actors show interest and put up substantial capital, may fund Government minority participation in a mortgage financing intermediary. In such a case, the Government and the Bank should probably avoid supporting such an intermediary for attracting funds in international capital markets through guarantee schemes; this support could undermine Morocco's motivation to develop the internal capital markets, and further put off, among others, the privatization of pension management. 83. Beyond the above immediate policies (para. 62), the road towards having a performing, mainly market based system to supply dwellings in about 10 years seems daunting. 34 Yet, we believe, the results will be overwhelmingly positive, and each step ahead will progressively be easier to take. On balance, that road is worth taking and of the Bank accompanying Morocco on such a road. c:\j\rnor\report\houstrat\report6.eng 11 42 November 29, 1995 Annex A-1 Page 1 of 2 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY The Urban Market Solutions: a b Where supply and demand meet: annual average for 1981-1990 Lots Houses Supplied by Pdoce Pice Number of Thousand of Number of Thousands of So/utions drhams Solutions drhams Public Agencies 23.500 c 6,900' purchased by: Low Income Households d 9,000 * 32,000 h 3,000i 150,000 - Middle-income households d 8,600 96,000 " 3,900 150,000 + High-income households d 4,700 g 180,000 h Private formal suppliers 7,300 k 41,900' purchased by: Middle-income households 8,100 " 200,000 - 500,000 High-income households 7,300 300,000 33,800 n 500,000 + Informal market 7,000 * 26,700" purchased by. Low-income households 3,500 18,000 Q 13.700 * 50,000 - 80,000' Middle-income households 3,500 18,000 4 13,000 P 100,000 - 120,00' Bidonville solutions 4,500' 4,500 purchased by: Low-income households 4,500 .4,500 5,000 + " Total urban lots supplied 42,300 Total dwellings supplied 80,000 Households moving to 9,000' another household Annual needs of urban 89,000' dwellings based on population growth Annex A-1 Page 2 of 2 The purpose of this table is to show what type of supplier provides dwellings to what economic strata, and at what price. b The figures are rough World Bank mission estimates, on the basis mainly of CIH' Etude du secteur de la construction (donn6es de base du secteur - d6cembre 1992) and Ministry of Housing's (MOH) Description et analyse du processus de production de logements au Maroc (1993 - Preliminary). C -Average as per page 8 of CIH's report, may be overestimated as the figure corresponds to an annual average of 1985-1990, whereas most figures in the table are annual averages over 1981-1990. d Based on Table 3, Volume 2 of World Bank report Kingdom of Morocco: Poverty, Adjustment, and Growth. Low income is the 27% of population with lowest per capital consumption, middle is 25% of population in the middle consumption range, and high is the 38% of population with highest per capital consumption. o Based on data in page 43 of MOH's report; estimated as an average of lots de recasement, plus one half the average of lots de restructuration. f Residual. Estimated at 20% of total, as officials in interviews mention that 80% of lots are given to population of low income, but the low-income demarcation is about the median income. Rough estimates assuming 80 m2 lots at DH 400/M2 for low-income households, 120 m2 lots at DH 800/M2 for middle-income households, 200 m2 lots at DH 900/m2 for high-income households, and 200 m2 lots at DH 1,500/M2 for privately supplied high-income households. From page 8 of CIH's. Economique in page 43 of MOH's report is assumed middle income; the balance low income. k Residual of estimate of total lots produced in the urban formal market, less those produced by public agencies; total lots was estimated on the basis of 2 dwellings per lot, a rounded ratio which approximates those cited in both reports. Based on CIH's report, page 7. Residual. Estimated on the basis of the above distribution of expenditures (0.38 x 89,000). o Assuming 2 dwellings per lot. Guess. * Assumed to be 5% of total urban, as 3% of total national is a commonly cited figure. Based on Caract6ristiques socio-6conomiques des m6nages des quartiers d'habitat clandestin, ministbre de I'Habitat, 1994. Based on Caract6ristiques socio-4conomiques des m6nages des bidonvilles, ministbre de I'Habitat, 1994. Annex A-2 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Trends in components of cost per m2 of floor area (in constant DH) 1972-80 1980-84 1984-90 1972-90 Major cities Land cost 9.7 -23.8 7.5 3.6 Construction cost 4.9 -4.3 2.1 1.9 Average cost per m of floor 6.5 -7.5 3.9 2.5 area Small centers Land cost 6.8 -20.2 9.7 6.2 Construction cost 4.9 -4.3 2.1 1.9 Average cost per m' of floor 6.5 -6.5 3.3 2.5 area Source: Urban Agency of Casablanca Annex A-3 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Supply of Dwellings for the Urban Poor a bc Average number of dwellings for 1981 - 1990 Number Percentage Public supply Lots only d 9,000 23 Houses 3,000 8 Subtotal 12,000 31 Informal supply 13,700 35 Bidonvilles 4,500 11 Forced Cohabitation C 9,000 23 Approximate Total 39,200 100 The purpose of the table is to highlight the extent to which the state covers the needs of the urban poor. Poor is defined as population in the third lowest level of consumption as per the World Bank's report Kingdom of Morocco: Poverty, Adjustment, and Growth. Arranged on the basis of table "The Urban Market Solutions". d More than one dwelling is generally built on these lots; owners often cede the right to construct a second floor to households who may not be poor. By definition, this solution does not produce new dwellings. PROCEDURES TO APPROVE MASTER URBAN DEVELOPMENT PLANS Central Government Local Government Department of Urban Planning Participation Initiate of the SDAU Cities & Communautes Elaboration2 Central Committee Elaborate draft Central Committee Local Committee Clearance Consultation Department of Urban Planning Modify draft p City 3 months Councils maximum Clearance Department of Urban Planning Prepare Decree Approval Visa MoF, MoPW, MoA, Mol* Decree Publication Official Gazette M MoF - Ministry of Finances MoPW - Ministry of Public Works MoA = Ministry of Agriculture Mol = Ministry of Interior m:\sylvie\urbi PROCEDURES TO APPROVE LAND USE PLANS Communes SDAU 3 months Department of Urban Planning Consultant prepares Revised Appraisal initiates the PA a PA draft Department of Urban Planning Draft PA Elaboration Public Inquiry Clearance of Revised City Councils Revised Ministries PA 2 months PA local services Consultation Visas of Department of U.P. Decree Decree MoF, MoPW* final document preparation signature Approval 0 MoF = Ministry of Finance MoPW = Ministry of Public Works x m:\sylvie\land PROCEDURES TO ISSUE LAND DEVELOPMENT PERMITS State City Promoters Central Local Aesthetic Road Appication Committee Commttee 11 [ mttee~e Clerce/ SoU 3 months max. If P.A. exists Authorization _earance_ Process f important projects Permit Drf Apol if differs from DoU DoU: If project on more 4 Approval than onecity * SoU = Service of Urban Planning (Local) 01 * DoU = Department of Urban Planning (Central) m:\sylvie\urb3 PROCEDURES TO ISSUE BUILDING PERMITS State Central Local Communes Individuals Other Ministries Aesthetic FRoad Local services Commee committee Authorization Flnitiate Local Office Process Processj of Urban Planning (SoU*) if Differs Building Approval from SoU* Permit DoU** Draft Approval SoU = Service of Urban Planning (Local) **DoU = Department of Urban Planning (Central) m:\sylvie\urb2 Annex A-8 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Land Distribution and Land Prices in Khemisset, Rabat and SaIN Khemisset Rabat Sale Distribution Price of Distribution Price of Distribution Price of of building undeveloped of building undeveloped of building undeveloped kind land land land land land Public 95 ha 50 Dh (a) 1285 ha 80/110 Dh 645 ha 80 Dh development 80% 69% Private 24 ha 200 Dh (b) 400 Dh 296 ha 300 Dh housing 20% 31% Illegal NA 45 Dh NA NA NA (c) 500 DH settlements Data sources: TUM Consultant Etude de faisabiliti du projet de resorption de l'Habitat insalubre de Khenisset TUM Consultant Etude de faisabiliti du projet Al Mohit a Laayada/Sal Direction of Housing and Division of Urban Planning in Rabat, Sal6 and Khemisset Provinces (interviews). (a) Hypothesis that 'Plateau d'Akraich' will be developed by public agencies (b) Small developments scattered throughout the city (c) Prices can reach 1000 Dh in well-situated lots Annex A-9 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Explicit and Implicit Subsidies In the Housing Sector ab (1992, unless otherwise specified) (DH million) Transfers to public developers 383 Interest subsidies 190 Income tax exemptions 116 Capital gains tax exemptions 564 Exemption of urban tax 24 Visible subsidies and tax exemptions .277 Tax expenditures in real estate profit taxes (average 797 estimate) e Possible value of deductible interest d 156 2,230e a/ The purpose of the table is to show Morocco's housing subsidies embedded in the fiscal system. b/ Arranged from tables in Housing Sector Strategy in Morocco: Visible and Less Visible Subsidies, David Sewell, September 1994. c/ Moderate estimate on the basis of the tax rate on interest income whose source is not declared. d/ Applying a 25% tax rate to DH 528 million deductible. e/ Technically, these subsidies are not additive; a total is given to provide perspective. Annex A-10 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Borrowers, and the Unsatisfied Housing Finance Market in 1992 a b Household with monthly revenues of No. of customers No. of urban % households with households no financing Less than DH 2,000 1,715 DH 2,000 - 3,000 197 } 27,000 93 DH 3,000 - 5,000 819 35,000 98 DH 5,000 - 10,000 More than 10,000 } 6,840- } 38,000 } 82 TOTAL 9,751 100,000 90 a/ The purpose of this table is to show the limited access to mortgage loans in Morocco. b/ Most of the information comes from CIH's Rapport d'activits de Vaercice 1992. c/ For the categories 5,000 - 10,000 and more than 10,000, it includes 1,119 and 573 houses respectively, financed by CIH and BCP. The figure 6,840 is a global estimate on the basis of the 1985 consumer survey, as mentioned in CI's report Etude du secteur de la construction - synth&se et simulations. Annex A-11 KHNGDOM OF MOROCCO HOUSUNG SECTOR STRATEGY Housing Finance in Morocco for 1991 and 1992 ab (CKH/BCP activities only) (Amount in DH million) 1992 1991 No. of No. of Anount Anowt No. of No. of Amount Amount loans dwellings approved disbursed loans dwellings approved disbursed Financing of developers 145 5,730 1,259 1.112 221 12,099 1,364 1,012 Mortgages General regime 2,281 2,730 471 525 5,124 5.873 825 744 of which: Purchase to developers 1.952 1,929 367 - 2,525 2.525 492 - Construction of 93 167 12 - 1,024 1,773 104 apartments Individual construction 236 634 93 - 1,575 1,575 229 - Special regime 1,497 1.497 105 75 1,167 1,167 86 89 Total Mortgages 3,192 3,641 527 600 6.291 7,040 911 840 Total Housing Lending 1,786 1.812 2,275 1,852 a/ The purpose of the table is to show housing financing activities in Morocco. It includes those of CIH and BCP. Financing of upper-income housing by banks is modest and covers higher-income households. b/ Information taken from CIH's Rapport d'activits de 1'evercice 1992. Annex A-12 Page 1 of 2 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Funds Flow of Contractual Savings in 1992 ab (DH million) DH million % SOURCES CDG System Pension and insurance reserves (long-term) 964 19 Social security funds 622 13 Postal savings and other 733 14 Increase in equity 284 2,603 51 Insurance System Mathematical reserves (long-term) 942 18 Technical reserves 1,403 27 Increase in equity 228 4 2,573 49 Total Sources = Total Uses 5,176 100 USES CDG System Govemment (384) -7 Housing finance through CIH 790 16 Financing local services through FEC 431 8 Other applications, mainly in affiliates 1.766 2,603 51 Insurance System Obligations from, or guaranteed by Government 546 10 Other obligations 566 11 Real estate investments 217 4 Stocks 564 11 Other uses __i 2,573 49 Comparison with banking system flows Sight and time deposits 10,921 Holdings of public debt 6,190 C Loans to private sector 7,142 d Annex A-12 Page 2 of 2 The purpose of this table is to show the importance of contractual savings (pension and insurance) in generating long-term resources relative to total generated resources and the size of the banking system. Prepared by mission, on the basis of CDG Rapport de gestion (1992), Compte-rendu de l'actualiti des entreprises d'assurances et de riassurances (1992) et Banque AI-Maghrib Annual Report (1992). Compensated by DH 5,168 million decrease in liquid assets and reserve accounts. d Excluding DH 473 million decrease in loans to specialized financial institutions. Annex A-1 3 Page 1 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Scheme of Readjustment Project Based on Sidi Abdellah Subdivision In Sale MAIN TOTAL AREA ROADS 100%=350Ha 15%= 52Ha AREA FOR DEVELOPMENT 85% 297Ha PUBLIC PRIVATE DEVELOPERS 70%= 208 Ha 30%=89Ha SALES TO RECOVER COST 21%=62Ha PUBLIC PUBLiC LOW-COST SERVICE PRIVATE HOUSING HOUSING-S 91%=271Ha 3%=8Ha c:J\mor\repoft\houstral\annexal3 Annex A-13 Page 2 COST BALANCE OF READJUSTMENT PROJECT IN SIDI ABDELLAH SUBDIVISION IN SALE Project area 350 Ha Price of Gross land (a) 300 Dh/m2 Price of Primary serviced land (a) 900 Dh/m2 Price of Serviced land (a) 1500 Dh/m2 LANDOWNERS PAY OFF Unit price Value (Dh/m2) (MDh) INPUT Raw land value 350 Ha 300 1,050 OUTPUT Primary serviced land 208 Ha 900 1,872 ADDED VALUE 822 PUBLIC DEVELOPER PAY OFF Unit price Value (Dh/m2) (MDh) COST Primary infrastructure (main roads) 350 Ha 150 525 Secondary and tertiary infrastructure 28 Ha 150 42 Administrative cost (7.5% of cost) 43 Financial cost (b) 37 BENEFIT Sale of land to recover cost 62 Ha 900 558 Sale of lots to low-income families (c) 19 Ha 600 91.2 Land for public services 8 Ha (d) NA NA (e) BALANCE 3 (a) Assuming no significant changes In private legal market prices. (b) Assuming Interest at 13 percent p.a., a one year average investment period, and linear disburse (c) 1,900 lots of 80 m2 each can be developed on these 19 Ha, assuming that secondary and tertiary roads cover 20% of the land. (d) Includes those publicly managed; others would be privatly driven. (e) On the long term, public developers should recover cost of land for public service areas. c:\\mor\repori\houstral\primary2.is Annex A-14 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Potential Reduction in the Cost of a House Current Cost 200,000 a Less: Effect of reducing urban standards b 18,000 Effect of lowering cement import duties 3,969 Effect of reducing concrete iron import duties d 3,373 Effect of reducing constructed area from 70 m2 to 60 m2 C 10,000 SUBTOTAL 35,342 Assumed combined effect of decrease in the price of raw land and 14,658 other efficiency gains Potential cost reductions 50,000 Potential cost 150,000 c:\j\tnor\report\houstrat\tables.eng a/ Present cost of a 70 m2 economic house in major cities, a generalized working assumption. b/ On the basis of 30% reduction in land cost (Francesco Bandarin's report, Annex 2, page 23), and a 30% ratio of land to housing costs. c/ Reduction of cement duties from 25% to 10%, assuming cement is 30% of structure costs, structure cost is 63% of construction cost (CIH study, p. 32), and construction cost 70% of total cost. d/ Reduction of concrete iron duties from 52.5% to 10%, assuming concrete iron is 9% of structure cost. e/ Assuming that a 10% reduction in area results in a 5% reduction in construction costs. EXAMPLES OF INSTITUTIONAL ORGANISATION FOR URBAN PLANNING FRANCE ITALY GERMANY ENGLAND COSTA RICA TUNISIA MOROCCO EVELS OF State, Region,depar- State.Region, Province Federal State,Laender, State, County,District State,Departement, State, Commune State, Commune OVERNMENTS tement,Commune Commune Commune Commune MN lRY IN Equipment Pubic Works Regional Planning Environment Oficina de Environment & Interior ClARGE Building and Urban Planilicadon Equipment development MASTER PLAN Schema Directeur Regional Plan Raumordnung und Structure Plan Plan Nacional de Schema Directeur Schema Directeur d'Amenagement et Landesplanung Desarollo Urbano d'Amenagement d'Amenagement d'Urbanisme (SDAU) et d'Urbanisme DECISION BY Commune Region Laender County National Agency State (M. of Environment) State (M. of Interior) LAND USE PLAN Plan d'occupation Piano Regolatore Bebauugsplan Local Plan Plan Regulador Plan d'Amenagement Plan d'Amenagement des sols Generale Urbain DECISION BY Commune Commune Commune District Commune State (M. of Equipment) Sate (M. of Interior) COVERAGE 80% of Communes 90% of Communes Widespread About 25% of NA 28% of urban 13 % ofcommunes the country communes BASIC INSTRUMENT Building permit Building permit Building permit Planning permission Building permit Building Permit Building Permit (Permis de construire) (Concessione edilizia) (Baugenehmigung) (Visado de Planos (Permis de Construire) (Permis de Construire) de construccion) DECISION BY Commune MunicIpal Municipal District Council Direction of Urban Commune Commune Building Commission Chief Planning Officer Planning Committee Planning of National Institute CONTROL AND State (M.of Equipment) Region Laender County National Institute State (M.of Equipment) State (M. of Interior) SUPERVISION (National Institute) (Regional Office) (D m:Uylvecps un Annex B Page 1 of 11 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY MEMOIR Jointly prepared by a Moroccan Interministerial committee and the Bank mission 1. From October 18 to November 1, 1994, as part of a follow-up mission, a World Bank delegation visited Morocco and worked jointly with a Moroccan interministerial commission to develop the basic components of a housing sector strategy. The Moroccan commission was composed of Messrs. Cherif Tahiri, Director, Department of Planning and Programs at the Ministry of Housing; Abdeliai Bousfiha, Director of Urban Planning and Architecture at the Directorate-General of Urban Planning, Architecture, and Land Development, Ministry of the Interior and Information; and Abdelghani Guezzar, who is in charge of real estate and business loans at the Ministry of Finance. The World Bank mission was composed of Mrs. Maryse Gautier and Messrs. Julio Linares, Henri Beenhakker, Robert Buckley, Francesco Banderin (consultant), Vincent Renard (consultant), and Franq6is Vaillancourt (consultant). This strategy should facilitate implementation by the Moroccan Government of directives issued by His Majesty that are designed to increase the low-income housing stock by increasing the role of the private sector and identifying suitable changes. For the World Bank, this implies that future intervention in the housing sector will be consistent with this strategy. This document provides both an overview of the problems facing the sector and appropriate policy recommendations designed to enable both parties to address those problems. HOUSING POLICIES 2. Background. In spite of the remarkable advances made in regulated urban development in recent years and the Government's efforts in the land and housing production sector, the housing stock falls short of demand (approximately 100,000 new urban housing units are needed for year 1993 to keep up with population growth and only 90,000 units, including 28,000 unauthorized units, are produced for the same period). Under the existing production system, the key component in this bottleneck is a shortage of land, which is exacerbated by dwindling reserves of public land. In recent years, land prices have risen at an alarming rate. At this time, the public sector dominates the production of building lots providing over 60% of the supply. This situation has caused distortions in the marketplace. The Team Maroc study shows that 40% of lots produced by the public sector are put back on the market by private intermediaries. These factors indicate that it is essential to reorganize the building lot production process to meet new challenges. Each component of this sub-sector, from regulated and functional urban development, to the instruments used in managing the real estate market, to infrastructure financing and the property tax system must interact in a consistent, effective manner. Such consistency can be achieved only by redefining the roles of the various urban development institutions. To achieve this, the role of private developers must be enhanced so as to increase the availability of land in the aggregate. The various components of this sub-sector are analyzed below. Annex B Page 2 of 11 SUPPLY CONSTRAINTS 3. Urban development plans. The linkages between the urban planning instruments - Sch6mas Directeurs d'Am6nagement et d'Urbanisme (SDAUs, or Master Urban Development Plans) and the Plans d'Aninagement (PAs, or Land Use Plans) - defined by the law of 1992 seem to be satisfactory. Similarly, the output of such plans meets requirements. At this time, 17 SDAUs covering Morocco's major metropolitan areas have been approved, and 217 PAs have been approved or are in the approval process. However, delays in preparing and approving the documents are an obstacle to lot production and urban development. This situation produces uncertainties both for land owners and real estate developers, and has contributed to instability in the market. From the quantitative standpoint, the land use plans generally provide for an adequate future supply of lots for urban development. Even so, from the standpoint of implementation, we noted the lack of timetables for developing and financing projects for land resource use and for implementing improvements, which could lead to spatial dispersion of development and under-utilization of infrastructure. The effective projects, mainly those initiated by public developers, are often carried out as a function of land availability, independently of their location relative to the land use plan. In implementing land use plans, we noted the absence of a market analysis and land survey component, which would optimize development options. 4. Urban planning standards. The standards specified by the urban development plans seem to meet international specifications routinely applied to regulated urban development projects, mainly in developed countries. Nevertheless, an analysis of various land subdivision projects indicated that the land use ratio for mixed housing areas was relatively low in comparison with the needs in a land shortage situation, which is the case at this time. The land use ratio varies from 35% to 45% depending on the project, whereas statistics on public and low-cost housing projects in other countries show that land use rates in the 55%-65% range can be attained. Similarly, while housing densities found in informal housing developments do not serve as a reference, they suggest a clear-cut need for more efficient land use. Differentiated standards for public housing projects and for slum upgrading have been adapted in the legislation and in practice. 5. Functional urban planning. There is little relationship between the urban development documents (SDAUs and PAs) and the location of urban development projects. To a large extent, the availability of land - primarily public land reserves which are being depleted, but also large, privately-owned tracts - determines the location of new developments. The PA, therefore, does not seem to be fulfilling its role, which is to serve as a framework for the development of building lots. Moreover, real estate development projects, whether public or private, do not fall within a legal and institutional framework for project definition and infrastructure financing. The review of the Zones d'Am6nagement Concert6 (ZACs or mixed public and private housing development areas) must therefore be conducted with a view to opening them to projects managed by private operators. These ZACs would supplement the instruments available to private developers of building lots, which has until now been limited to housing development. Annex B Page 3 of 11 6. Public supply of serviced lots. As discussed during the World Bank's previous mission, the public developers (SNEC, ANHI, ERAC) are unable adequately to meet their goals in the area of equalization. They have not been able to deliver sufficient numbers of serviced lots intended primarily for low-income housing, or install enough core infrastructures to enable private developers to increase the supply of affordable lots. The depletion of public land reserves and the mounting problems associated with expropriation make it more difficult to apply the equalization system and would require increased Government subsidies to produce the same output, as Team Maroc underscores in its study. 7. Managing the land market. Under prevailing conditions, the operation of the land market lacks transparency. While information on land prices is available locally, there is no system for consistently monitoring market conditions. Extensive under-reporting - for obvious tax-related reasons - of prices declared when recording land transactions contributes to lack of transparency. Publication of transactions at very high prices could contribute to disseminating the notion that prices are abnormally high relative to real market conditions, while concurrently encouraging a propensity among landowners to withhold their land from the market. It should also be emphasized that there are many complex land ownership situations, particularly those involving successions, which further restrict market fluidity and land supply. 8. Land taxes. Taxes and duties assessed on land and real property (urban tax, council tax, tax on undeveloped lots, business license fees) all have similar shortcomings: They are expensive to administer, they are relatively unproductive and, most importantly, they are inconsistent with urban development objectives. This is particularly the case for the tax on undeveloped lots, which was reinstated to serve as an incentive, but which plays only a marginal role because of its low rate and its limited scope of application. There is not yet any form of taxation on land situated outside urban areas. Hence, it costs nothing to "store" such land from a tax standpoint, even though its location in urban fringes and the proximity of infrastructure substantially increase its market value. In this respect, we note that the tax to be collected by local governments to recover the value added by public utilities is not applicable. Finally, we underscore that tax assessments based on rental values are inadequate, since reassessments take place only every five years and miy not reflect fair market values, thereby introducing distortions among regions and taxpayers. 9. Constraints on the supply of rental housing. The regulatory framework for rental units is an important factor in restricting the supply of housing. Under the current law, delinquent tenants can be evicted only by following the customary civil procedure, which is both cumbersome and time consuming. Moreover, this law specifies that rents can be increased only every three years, and only after filing a petition with the court if the tenant objects. This is one of the main reasons why investors are reluctant to invest in rental property or to rent dwellings that were built for sale. Hence, several thousand urban dwellings.now stand vacant because their owners believe that renting them to tenants who may not be creditworthy could prevent them from recouping their investment by selling the units. A bill now before Parliament proposes to solve this problem by enacting more efficient procedures for evicting delinquent tenants. Annex B Page 4 of 11 However, it maintains the legal procedure for raising rents every three years. It would appear that this escalation provision will remain in effect until such time as streamlined eviction procedures and land market liberalization have considerably increased the supply of housing, particularly for low-income households. Moreover, it should be noted that Sophal/CIFM, an affiliate of the Caisse de D6p6t et Gestion (CDG), manages about 54,000 public housing units that are rented at rates substantially below market. 10. Emport duties. Another factor which affects the supply of housing is import duties. For example, the ad valorem tariff on cement is 25%, which results in a commensurate increase in the price of this commodity. Since there are several cement factories in Morocco, and cement should be produced efficiently in all countries of intermediate size, this import duty and the attendant high price of cement results in continuing inefficient production of cement, or in generating monopolistic profits, at the expense of people who need housing. Similarly, the import tariff on concrete reinforcing rods is 50%. According to the Team Maroc Study, from which this information was abstracted, the only factory of any significant size which makes reinforcing rods is operating far below capacity. This implies that the import duty protects inefficient local production and also increases price paid by the end users. If tariffs on these two products were lowered to levels in line with those applied in countries where foreign trade has been liberalized to a large extent, the price of a DH 200,000 dwelling would be cut by about 10%. Other import duties are also high and probably increase housing costs substantially. 11. Other fiscal measures. Other taxes restrict the supply of housing, while direct or indirect fiscal subsidies have either failed to increase supply, or have resulted in higher demand from groups other than low-income categories. Because these potential buyers face a monopolistic supply situation from real estate developers, the tax deduction for real estate profits from (see Morocco Housing Sector Strategy, overt and hidden subsidies) results in higher profits for private real estate developers rather than an increase in the supply of building lots and housing. In addition, the deductibility of mortgage interest payments from gross income is more advantageous to wealthier homeowners than to lower-income groups. In Morocco, as in all developing countries, the tax deduction for mortgage interest is far more regressive than in developed countries. In developing countries, low-income groups do not have ready access to the mortgage market, and this deduction, which benefits the affluent and middle classes, reduces the revenues available for direct assistance to the poor. In developed countries, this deduction mainly results in over- consumption of housing, which benefits the real estate sector. Annex B Page 5 of 11 CONSTRAINTS ON DEMAND 12. Demand for housing. It is generally recognized that a lack of long-term financing reduces demand for housing, particularly from modest-income groups. Nearly 90% of households finance their home purchases from their own resources or rely on informal sources of financing. Over half of urban households who purchase a building lot obtain financing for the total cost, which is set by public land development agencies, generally at below-market prices, minus a down payment. 13. Constraints on the securities market. Financial and institutional constraints on the financial market, which the Government has begun to address, continue to impact the availability of mortgage financing at market rates. Moreover, CDG holds a monopoly over the administration of the social security and pension funds, as well as postal savings accounts, and enjoys an explicit guarantee from the Government when investing in securities issued by government agencies. CDG's monopoly status facilitates financing of other government agencies. The Government's explicit guarantee, in addition to its implicit guarantee, means that funds that CDG could have made available to mortgage lenders in an unregulated market are not onlent to these lenders. Even if there existed a supply of mortgage-backed securities issued by lenders - which is not the case at present - the requirement that insurance companies keep 40% of their reserves in the form of government-guaranteed instruments would sharply curtail their investment in such securities. 14. Constraints on the supply of mortgage loans. In addition, a number of policies limit the advantage to banks of offering mortgage loans. Such loans are subject to an interest rate ceiling, a requirement that is expected to be lifted in the near future. Only the Credit Immobilier et Hotelier (CIH), a government mortgage lender, has the right to foreclose promptly on borrowers whose payments are in arrears. The other banks must follow the lengthy procedures provided under the Code of Civil Procedure. As for interest discounts, these are not available to all banks, and they may not be as effective in reducing the risk of default as a rebate paid at the time of the purchase and applied toward the borrowers' down payments. 15. Financial liberalization. The financial liberalization program which will be completed over the next few months will enable the establishment of an effective mortgage market. We note that only through this liberalization, conducted within the parameters discussed by the Ministry of Finance and the World Bank, will it be possible to increase housing finance in Morocco. Annex B Page 6 of 11 HOUSING STRATEGY 16. Context. To develop a housing strategy for the intermediate term, the first thing to be taken into consideration are the basic priorities; the second is to examine the prospects for the housing sector over the next 15-20 years. Among the priorities, those designed to eliminate constraints on the supply of urban building lots rank much higher than those designed to stimulate demand. Raising subsidies and facilitating loans without altering the other parameters would not be sufficient to increase land supplies to any significant extent. We have identified the following components of a long-term strategy: (a) Local governments of urban metropolitan areas and their agencies would be responsible for functional urban planning and for the delivery of serviced land provided with core infrastructure. They would recover their costs and generate a profit by selling improved lots; (b) The private sector would provide an unrestricted supply of building lots and housing to middle- and upper-income households; (c) The Government would provide transparent subsidies to low-income households and would rely on the private sector for construction whenever possible. In the intermediate term, specific actions should be adopted to improve this sector without erecting barriers to long-term development. Overall, the strategy for the intermediate term should therefore provide access to decent housing for households in competitive markets and as a function of income, and should enable the public sector to stimulate the supply of housing for needy households with a minimum outlay of public funds. This is the context underlying the basic strategy proposed. Our recommendations are outlined in the following paragraphs. 17. Regulated urban planning. The importance of urban planning in the land delivery system warrants the adoption of a number of recommendations: (a) Accelerating the review and approval of urban development documents; (b) Preparing an implementation timetable that differentiates areas to be developed in the short term from those to be developed in the future; (c) Combining the land survey with preparation of the development plan so as to focus urban development options on more appropriate areas and more suitable instruments. 18. Urban development standards. The effective use of land resources is largely dependent on urban development standards. In this area, we recommend: (a) More efficient land use for developments designed for public and low-cost housing projects by raising the land use coefficient to 50% . exclusively in residential areas and to adopt a land occupancy factor of about 2; and (b) Planning for differentiated development to make property ownership more affordable and to amortize infrastructure costs over time. Annex B Page 7 of 11 19. Operational urban development instruments. Under prevailing conditions, it would appear necessary to design and to apply the operational urban development instruments described below, which entail interaction between the government and the private sector: (a) The establishment of mixed housing development areas (ZAC) for major projects where the land component is controlled, with negotiations between private developers and the Government to define not only the physical features of the project but also the allocation of responsibility between the two partners for financing infrastructures. (b) Land consolidation-development, a government measure that is more appropriate than voluntary consolidation in situations where zoning is fragmented or for the development of large tracts of land; and (c) In both cases, a mandatory system should be instituted for the free transfer of land easements (perhaps on the order of 30%) for the construction of infrastructure. Moreover, the local government could recover the cost of core infrastructure by selling plots, with the profits to be used to provide housing for low-income households. 20. Management of the land market: instruments for intervention. A system should be instituted under which the Government could exercise preemptive rights in areas to be restructured and in future urban development areas. In this respect, whenever the Government creates an area that is subject to rights of preemption, it should concurrently designate a "reference price" based on market value as of the date the area is established, but not including prospective rises in connection with the announced land development project (and the attendant off-site services to be installed). 21. Information tools. To address the relative lack of transparency in land markets, in the initial stage, a simple system should be developed to monitor land values based on appraisals by professionals in the sector (government agencies, real estate agents, private and public real estate developers), and to produce charts of the reported land prices. These charts should be widely distributed and used to check prices applied in expropriations, and particularly to check the prices of transactions, both for "adoulaire" contracts and sales of registered plots. In the second stage, publication of all real estate transfers and land transactions could be envisaged. The main goal of this measure would be to enhance the transparency of markets. 22. Infrastructure financing. The provision of infrastructure -an essential component of the land delivery system - requires major reforms in the short term: Annex B Page 8 of 11 (a) Measures should be adopted to enable developers, including those working in the private sector, to provide infrastructure under competitive conditions, under the supervision of government authorities; and (b) A strategy for financing core infrastructure should be defined based on an operational instrument such as a ZAC or subdivision. 23. Mission of public operators. In the context described in paras 19-22 above, it is essential to redefine the mission of the public developers (SNEC, ANHI, ERAC) by eliminating equalization as a system for producing lots for both low-income groups and wealthier population categories. The mission of these developers would be to increase the supply of building lots in the following ways: (a) On public land, or on tracts obtained quickly and without expropriation, by installing core infrastructure, selling large tracts to private land developers, and setting aside a percentage of such land for low-income housing; (b) For tracts owned by a small number of owners, by establishing mixed housing development areas (ZACs); and (c) For larger tracts, or if there is a large number of owners, by initiating land consolidation-development procedures. In each urban areas, such projects should be part of an overall strategy designed to reduce land costs, which requires intervention in a large land area by a single developer. 24. Land taxes. With a view to establishing an annual property tax assessed on the market value of the land and to serve as an incentive, it will be necessary to devise a method for assessing land prices based on deeds of transfer. The existing components of the land tax system would have to be completely overhauled and a new system erected on the following principles: (a) An annual tax that would apply to all land with the exception of farm land, payable by the landowner, assessed on the basis of market value should be established. The tax basis should be reassessed regularly. The tax rate should be defined at the local level. It is important for the tax to be based on market value alone and not on the zoning under the land-use plan. This land tax would supplant the cumbersome and ineffective tax on undeveloped land, which was reintroduced in 1989; and (b) In order for the tax on real estate profits, which is a sound principle, to function properly, coordination must be developed between the tax collection agencies, mainly the land record offices, and the communes (or the agencies). This tax Annex B Page 9 of 11 should be a local tax and would require that the land record offices systematically notify the communes of transactions that have taken place in their districts. 25. Other fiscal measures. These would include: (a) Lowering import duties, at least for cement and reinforcing rods, to not more than 10-15% to reduce building costs; and (b) Abolishing the tax deduction for real estate profits (hence, abolition of the real estate code) and for mortgage interest payments, so as to increase the resources available to the Government to subsidize the poorest population groups in a transparent manner. 26. The rental market. In the rental market, we would recommend the following: (a) Enacting the law pertaining to payment of rents; (b) Selling public rental units by taking advantage of rent increases and therefore of higher property values resulting from enactment of the above law; and (c) Conditions permitting, the enactment of a law authorizing rents to be determined by mutual agreement between landlord and tenant and abrogating the requirement to file with the court for authorization to increase rents every three years. 27. Financial aspects. In the financial arena, the following measures would be needed to enable to mortgage lenders to secure access to long-term funds: (a) Develop a market for mortgage securities, similar to that proposed by the Commission that studied the reform under the auspices of the Ministry of Finance and Banque AI-Maghrib (the Commission's proposals were summarized in one working paper that specifies certain adjustments, and the major problems in the real estate market were addressed in another). According to this proposal, the banks would issue securities that would be guaranteed by a portfolio of mortgages. Notwithstanding legal or practical problems associated with the mortgages, we do not think it would be either necessary or appropriate to create an institution that would buy the mortgages from the original lenders and issue securities to institutions with long-term sources of capital. To minimize supply constraints, a modest increase in funds available for mortgage loans would be a success. The transfer of mortgages and the concomitant elimination of all risks to the originating banks would either place a heavy risk burden on an institution acting as an intermediary between the banks and the long-term fund providers, or entail excessive administrative costs associated with risk management. However, unlike the Commission, we do not think the Government should Annex B Page 10 of 11 provide any guarantee for this market, because this indirect assistance would create distortions and the potential risk that the Government would ultimately suffer the consequences of poor-quality loans made by banks which would not be assuming the concomitant risks. On the other hand, we agree with the Commission in that we believe there is no place for an intermediary in a mortgage-backed securities system; and (b) Withdraw the Government's explicit guarantee on securities issued for CDG by an organization, whether public or private. The purpose of this measure would be to enhance CDG's sensitivity to interest rate differentials, and it may provide access to CDG's funds for mortgages issued by the banks. The foregoing measures could and should be implemented forthwith. 28. - Foreclosure rights. To encourage banks to increase their production of mortgage loans, the right to foreclose on property assets, which is reserved exclusively to CIH at this time, should be extended to these lenders. 29. Financial assistance to individuals. Two other measures should be taken into consideration in the future. The first would be to establish a small-scale home purchase savings system (providing this clearly is warranted), i.e., a contractual savings system guaranteeing that depositors will be eligible for a loan after they have accumulated a certain level of savings over a given period of time. Such a system would help modest-income households to secure mortgage financing, but there is a risk that it could generate negative cash flow. The second measure would be a system of Government grants to be applied to the initial down payment on a home purchase. This system, which has been successful in Colombia and Chile, lowers the minimum income needed to qualify for a mortgage loan, thereby reducing government intervention in the provision of public housing. These two measures should be examined more comprehensively, and should not be implemented until the supply system has adjusted to the liberalization of markets and housing. 30. Assistance to low-income groups. While the operational urban development instruments designed to increase the supply of land will help to provide building lots and, potentially, housing for the poorest households, the financial reforms will be of no use to these groups. This is inevitable, since any attempt to require that financial institutions and private entrepreneurs assume unwanted risks will fail and will cause other aspects of these reforms to fail as well. Hence, two additional measures requiring in-depth examination are recommended: (a) Implementation of a system of direct assistance to very low-income households who would be identified by local governments as eligible for rent subsidies to be paid directly to the lender and financed jointly by the local government and the Treasury; Annex B Page 11 of 11 (b) Establishment of a similar program for low-income households who are building houses in areas that are being developed progressively, either under reorganization or in the process of being upgraded. 31. Immediate actions. The measures that offer the best combination to ease the implementation, importance, and urgency, and that should therefore adopted as soon as possible are the following: (a) Instituting a mandatory system for the free transfer of land easements to recover infrastructure costs and provide housing for low-income households (para. 19(c)); (b) Refocusing public developers toward the delivery of serviced lots provided with infrastructure, initially by selling large tracts of land to private developers while setting aside a percentage of the land for building lots for low-income families (para. 23); (c) Lowering tariffs on cement and reinforcing rods (para. 25); (d) Enacting the law pertaining to payment of rents (para. 26(a)); (e) Eliminating the Government's explicit guarantee for securities issued on CDG by public or private organizations (para. 27(b)); and (f) Extending the foreclosure privileges to all banks now reserved to CIH. 32. Additional studies. While some of the recommendations outlined above could be implemented in the short term after an assessment conducted by the responsible parties, others require considerable in-depth study. Studies to be conducted with the assistance of the World Bank could be refocused so as to avoid a reexamination of policies that have already been evaluated and to concentrate on a more comprehensive review of problems relating to implementation. The authorities and members of the World Bank will hold discussions on the specific changes in the terms of reference for these studies. m:\sylvie\joint-am.eng c:\j\mor\report\houstrat\annex-b December 20. 1994 Annex C Page 1 of 31 KINGDOM OF MORROCO HOUSING SECTOR STRATEGY Report on the impact of urban planning and regulations on land use efficiency by Francesco Bandarin, consultant Summary and recommendations 1. Urban land supply and current planning practice in Morocco. 2. Urban planning and land supply 3. The impact of planning regulations and standards 4. The impact of subdivisions on land supply 5. An exercise in land use efficiency Venice, December 1994 Annex C Page 2 of 31 Summary and recommendations This report analyses urban land use planning practice in Morocco, with the aim of evaluating its impact on land use efficiency. The main aspects of the land development process examined by the report are : a) the system of land use planning tools and their influence on land supply. Land planning in Morocco is essentially based on a general Structure Plan (SDAU) for major cities and a more widespread local Master Plan (PA). Both SDAU and PA establish the total amount of land that can be developed and the use densities. In aggregate terms the total theoretical supply of land envisaged by the plans responds to the needs generated by urban growth. The planning system, however, requires streamlining, since the considerable delays encountered in planning preparation and approval processes generate uncertainty and disturbances in the land market. b) the system of land use standards and regulations. Standards used in land development in Morocco are not substantially different from those in European and other developed countries. This may be acceptable for middle and high quality land developments, but becomes a restraint in the case of low-cost housing. The reserve of large amounts of land required to meet urban standards increases the scarcity and cost of land without really producing an effective benefit, since there are huge delays in introducing services. The ongoing rehabilitation processes of illegal housing reveal that official policies have already changed and now accept the application of differential standards for neighborhoods with different economic levels. c) actual land development practice. This analysis has been applied to a number of large publicly sponsored projects currently underway in order to evaluate the actual impact of design practices on land price and land availability. The analysis shows that public land subdivisions have lower density levels than allowed by the official land use plans and that the criteria used in the design do not optimize land use density. This is mainly due to the excess size of streets and public spaces, and to the flat application of official standards. The result is that less than 40% of land is effectively used for residential purposes. The main conclusions of this report are: a) A reform of the planning system is required to streamline the drafting, approval and implementation procedures. Key elements of a reform strategy are: 1) the streamlining of drafting, approval and revision procedures of the Plans d'Amdnagement; 2) the decentralisation to the Wilayas of planning approval powers, while central government only retains a supervisory role. b) Standards and regulations should be revised to leave greater flexibility for developers in the choice of housing types. The leading criteria for the adaptation of standards should be the increase of land use efficiency, i.e. density and land coverage. This can be achieved by adjusting some of the prescribed standards and by allowing higher density housing types to be built. Furthermore, differential standards for infrastructures and services should be introduced not only for rehabilitation projects, but also for new low-cost housing schemes. Annex C Page 3 of 31 c) Project design criteria, especially for low-cost housing land subdivisions, should be reoriented to achieve higher land use efficiency. In particular, projects should aim to reduce the amount of land used for circulation, parking, urban services and community facilities, by using more appropriate design criteria than those allowed by a flat application of official standards. A more efficient use of land zoned for residential purposes and included in large public subdivision plans can accommodate up to 50% more housing units than current plans. 1. Urban land supply and current planning practice in Morocco. Urban planning and regulations have a strong influence on land supply and consequently also on land prices. Although planning should not be considered the only or principal factor of land price increase, zoning and standards undoubtedly generate significant variations in the price of land by altering the type of development which the market might choose as optimal for any specific land plot. The extremely high land price levels reached in many Moroccan cities in recent years suggest a closer look should be taken at the effects of planning on land supply. Recent Ministry of Habitat data on housing and land prices effectively highlight the size and the growing weight of land costs on the total housing production cost. The following table summarizes these findings, which show how land cost now represents a significant share of the housing production cost, even in situations favored by the presence of large public land banks. Table 1: Land costs for different housing types TypeDh/m2 built surface % Construction % Land Low-cost(Moroccan housing) 1,628 75 25 High rise 2,340 74 26 Villas 2,730 69 31 Source: Team-Maroc, 1994 As most of the public land has now been taken up, and the prospects of reconstituting the land banks are uncertain, land prices and their incidence on housing are bound to rise sharply. This trend is already observable in the major cities of Morocco, where land market tensions are more pronounced. The following table illustrates the price level reached in several Moroccan cities in recent years. The incidence of land prices on built housing ranges from 30% in the average size cities to over 40% in Casablanca. Annex C Page 4 of 31 Table 2: Land costs for different housing types crry POPULATION COST OF LAND COST OF TOTAL % LAND 1982 Dh/M2 CONSTRUCTION Ain Chock (Casa) 2,158,350 1,185 1,859 3,044 38.93 Sidi Bernoussi (Casa) 2,158,350 1,143 1,530 2,673 42.76 Hassan (Rabat) 518,616 880 1,365 2,245 39.20 Tangier 266,346 616 1,365 1,981 31.10 Agadir 110,479 578 1,160 1,738 33.26 Said 289,391 475 1,115 1,590 29.87 Fs 448,8231 415 1,020 1,435 28.92 Mekn4s 439,7281 3871 1,090 1,477 26.20 ource: Lehzarn, 1994 In Morocco planning greatly affects the public land production system, which is responsible for about 60% of total land developed in the formal sector. This type of land production is characterized by large development operations often with correspondingly large public land banks. In recent years, as the public land banks gradually became exhausted, urban planning schemes have reflected the new model of land development pursued by the public authorities, i.e. the creation of large new urban perimeters to be progressively expropriated by one of the major public land developers (SNEC, ERAC, ANHI). In this way, the plans set aside large tracts of land for future development which, because of the difficulty and cost of the expropriation process, are defacto excluded from the market. This is one of the main reasons for the current land shortage, and the consequent price increase. But planning also affects private land development. Private development in Morocco is only about 40% of the total and is specialized in small subdivisions, usually in the medium and high quality range. The existence and effective validity of plans, their zoning and density prescriptions, as well as the regulations and standards applied, are factors which significantly affect private investment decisions and the economic and financial feasibility of land development projects. All of these factors thus affect land supply and price. Finally, planning influences the process of illegal land development both directly and indirectly. Since restrictions in land supply and regulations increase the supply and price of land, a growing number of people are driven to develop land illegally. As will be shown below, this type of illegal development provides great cost advantages to the promoters, but transfers a significant part of the cost of infrastructures to local government. Annex C Page 5 of 31 All three main systems of land development - the public, the formal private and the illegal private systems - are affected by the planning system. In recent years, these influences merged with shortages of land supply and growing market tensions which pushed prices to levels universally judged as intolerable for the effective development of low-cost housing policies. The planning system therefore requires re-examining and reforming. Accordingly, this report will examine the impact of land planning and regulations on land supply, focusing on the following three aspects: a. the urban Structure Plans (SDAU) and Master Plans (PAs) system b. urban land development regulations and standards; c. the practice of subdivision planning as it is actually implemented. 2. Urban planning and land supply. The basic urban planning tools in Morocco are the Schema Directeur d'Amenagement et d'Urbanisme (SDAU) and the Plan d'Am6nagement (PA). SDAU is essentially a Structure Plan defining the long term development strategies in land use, transportation, services and industrial development. With the exception of Casablanca, the SDAU has no binding powers on private rights. PA is a Master Plan: in accordance with the provisions of SDAU, it defines the detailed zoning of the city, and indicates the uses, standards, and procedures for land development. This plan has binding powers on private rights. PA is the basic tool of urban development management, but large developments are regulated by Subdivision Plans (Lotissements), which define the plots of land and their uses and provide a detailed scheme of the new physical transformation, including streets, urban services and other land uses. A relatively complex plan, the SDAU has only been used in the major cities of the country (only 7 SDAUs have been approved to date, while 10 more have been prepared), whereas the PA is widely used as basic planning tool (so far 136 Plans have been approved, 81 are awaiting approval, and a further 146 are being drafted). These data indicate that the general planning system established in Morocco is relatively well developed, and its structure affects the availability and price of urban land supply by a) setting the amount of land to be developed; b) establishing the use density of the new developments; c) defining the amount of urban services to be built. Annex C Page 6 of 31 There are not enough figures available to make a global evaluation of the impact of local plans on the market. Nonetheless, data on the major SDAU, covering the most important cities and comprising over 60% of the urban population of the country, reveal that at this level of planning sufficient land is made available for development. The following table summarizes the data: land available for future developments nation-wide amounts to over 63,000 hectares, which is enough to meet the housing requirement for at least the next decade. Table 3. Land availability of towns with a SDAU City Urban perimeter Built land Buildable land Casablanca 14,000 8,500 13,000 Rabat 11,550 8,800 10,925 Other cities 57,223 25,390 39,151 Total Morocco 82,773 42,690 63,076 Source: Data supplied by Ministere de Imabitat, 1993 By applying the low range of urban population densities encountered in Morocco (200 to 300 inhabitants/hectare ) to this figure, the amount of land available would allow for the settlement of a population from 12 to 19 million people in the major urban centers of the country, or for around 2.5 to 3.8 million housing units. This figure is higher than the total estimated national housing requirement for the same period (the estimates vary, but most authors agree that the cumulated need in the year 2005 will be at least 2 million housing units. See Lehzam, 1994). Similar conclusions may be drawn for all the cities considered . For example, the SDAU of Casablanca allows for the development of 13,000 hectares of land, which would be sufficient to house an additional population of 2.6 to 3.9 million people (table 4). This logic also applies to the PAs, which have a shorter duration than the SDAUs. The SalM Plan d'Amdnagement, for example, allows for the development of 1,622 hectares of land. Even assuming that a large share (500 hectares) consists of existing illegal settlements, this leaves over 1,100 hectare of land for new developments. At the current average density in Sal6 (488 inhabitants/hectare), this would mean the settlement of over half a million people. TABLE 4. BASIC INFORMATION ON URBAN PLANNING IN RABAT-SALE AND CASABLANCA MASTER PLANS (SDAU) AND LAND USE PLANS (PA) OF RABAT AND CASABLANCA. BASIC INFORMATION I_i RABAT-SALE METROPOLITAN AREA NOTES SALE TEMARA RABAT TOTAL CASABLANCA Land area inside the Master Plan (ha) 1 79,453 49,310 11,751 140,514 9,800 d area inside the Land Use Plans (ha) 2 6,395 2,078 7,237 15,710 Residential area of the Land Use Plans (ha) 3 2,558 831 2,900 6,289 21,000 Built residential land area of the Land Use Plans (ha 4 936 376 2,200 3,512 otal built land (ha) 2,340 940 5,500 8,780 8,000 Population inside the Land Use Plans area 5 401,800 124,200 625,400 1,151,400 2,300,000 (1) Figures of 1990 (2) Sum of developed land area in 1987 and of land to be developed after 2007 (3) 40% of the area covered by the Land Use Plan (4) 40% of the residential area of the Land Use Plan (5) Figures of 1987 m _j 0 : Annex C Page 8 of 31 In conclusion, if the figures are considered in an aggregate way, in quantitative terms no effective restriction to land availability may be attributed to the Structure Plans (SDAU) or even to the Master Plans (PA). Needless to say, however, these calculations are purely theoretical. Actual land availability for development depends on many other factors, both physical (distance, technical factors) and juridical (tenure, ownership, etc.). An effective analysis of the impact of regulations on the land market can therefore only be made in terms of project implementation. But at this level of planning one vital aspect should not be underestimated: the impact on the land market of plans during their preparatory and approval phase. At such times uncertainty may be introduced to the markets as operators hesitate in evaluating potential developments. These phases should therefore be reduced to the strictly minimum time required for the preparation and the approval of the plan. In Morocco, however, both the preparatory and the approval phases tend to be very protracted, thus increasing market uncertainty. The weakness in the approval process directly affects the implementation schedule for plans. This in turn constitutes the most severe obstacle to national government efforts to control urban growth in recent years. Dilatory approval processes are accompanied by a lack of effective phasing in the implementation of the plans, and by a failure to establish any real linkage with the investment budgets of central and local government. No recent evaluation of the effectiveness of Moroccan urban planning is available. The results of previous analyses (Keene, 1985), however, and of more recent case material (Bahi and Alami, 1992) confirm that implementation of urban plans has been far from satisfactory. Today urban planning provisions are essentially made for the purposes of selecting sites for large-scale public land developments. They also enable ordinary urban development management practices (development permits and subdivisions) to be conducted, but have revealed serious weaknesses in ensuring that the prescribed level of services and infrastructure standards are observed. An evaluation of the case of the 1981 SDAU for Tangiers, for example, reveals that after almost a decade of the plan, only 6 out of the 39 planned primary schools had been completed, and only 7 out of 15 high schools, while only 1 out of 3 hospitals was under construction and none of the 10 Dispensaires or the 6 Health Centers had been built (Bahi and Alami, 1992). This situation cannot be directly ascribed to the plans, since an important role is also played by the availability of financial resources for investments. What does happen, however, is that large Annex C Page 9 of 31 amounts of land are tied up, since their use is restricted to services prescribed by the Plans. These, in turn, will probably never be built according to the planned standards. The complex preparation, approval, and implementation procedures are equally responsible for cumulative delays that tend to make the plans outdated even before they are actually operational. In most cases, the SDAUs prepared in the 1980's are due to expire in the next decade. This problem is even more acute for Master Plans, which are only valid for 10 years. All of these considerations should induce the national authorities to restructure the present planning system by introducing streamlining and flexibility. For instance, two positive measures would be to unify the approval procedures for the plans and to set limits to the length and extent of consultations. This would greatly accelerate the preparation and approval of the Plans d'Am6nagement, which should also be more carefully phased according to the effective availability of financial resources. 3. The impact of planning regulations and standards Planning and regulations may directly affect land supply by restricting the use density of land. In practice, for a given housing development project, standards affect density by establishing the amount of land which is actually saleable, as opposed to land for public infrastructure (streets and parks, and other community services). Furthermore, standards determine the amount of land actually used for residential purposes, as compared to land used for facilities and other non-residential uses. Urban standards applied to land development projects in Morocco are usually included in the Master Plans (PA), and are specific to the zoning area of the city and to the type of housing development allowed. In general terms, the standards used in Morocco (see Tables 5 and 6) are not substantially different from those in many European countries. Even if a direct comparison of these standards with the practice of other countries is greatly complicated by the decentralization of urban land development regulations to local government, the Moroccan standards generally reflect the accepted international practice of land development. But standards can only be truly evaluated in a local context. From this point of view, standards applied in Morocco can only be considered a suitable guideline for the design of medium and high-level housing development, but they are certainly not appropriate for the low-cost housing and resettlement projects presently underway in Morocco. An overall evaluation of the effects of land development standards has not yet been conducted in Morocco. Despite the lack of thorough data, a recent Team-Maroc evaluation financed by the Ministry of Housing (Team-Maroc, 1994) points to the need for a re-evaluation of current infrastructure and land use standards in order to cut final costs. Annex C TABLE G. COMPASON OF URBAN STANDARDS nN mOOCCo AND OTå TE CMARCNOF ST AN~DARS AMO@ DFR T COUVNTfRIES._________ ZIZZ I TU-In ZZZUEI.ZZER g 1N CASA ___RAAY- i Tb~~ ItGYP.R _ JPLo / AREA RATIo 90_ma__ t_:D___ m=n3 ow4n ct _- cow .co 1 1.0 ,F1 W AM) AR A Yt 1.94 O ___-4¯¯_ '3.0 0.co ¯_.0b__ e S ____ 1. MAIL L-lAMb~avimãz mfð¯ - --- -- -@ - i- ¥n -~¯¯0 -_- I m0 DYN~~ ~ 2 . 9m @10 |Sa WMIkMAL LOT a_25 WDEMDUAL MOUOING i________ ___2 2 _2___2__c___ _ _ _ _flE musi 2a ¯ - ~- WA ¯D-CD m _0 20 m ¥ Jo '7 I ac [ eo MA____A __B __t______E___Y___-20 1-.17 m 25 m ~____ ___12 __ ___ c_ _ _F_¯© ____AREA__AY____.2¯-0. 1.0¯ T .51ä 0&. | ____.2 PF A _1- _ . 0.04. [ 1.2 G NIMLLO SIZE ._0-2__204 _2_Q _ _2_-_ _ co__c byIøIyAL LOT WMT - _ _ ~_M e m -im wJw: c |MA.IMAL UILDId HEIGHT[ 1 1 mm i§ lam 12f RATO - -i _ 22 _ _ - _ .0 3.0_ _ MAXMA~ LAND WOVE.RAGE RATIO I - 0% IEM cn &o 0 I WHIALLiPT VVIDII m GMm l r Ifim ______ K____D _ MlA~A~ BUILWIMG PEIGHli i~ 0D-12 m loff l lm lam [l m IZ j (2)18 mrone dagaO~2 co tno sto me:n sY?sso 2900 inh2 o.) nd or cPmInd (co.oinhab.). 10m~ h~ ~~b~~ Iuc ~t aih ý= ~~ ~p (3) ataly dgcs dosr c3kmb"ieIed ~ y, din to Iho C ~ide" ~ coad b Iho m eg0ak . Doto oo St&n Qsm Ito d ola ;: (ow LtVbo FdsU-Votahb =i Pc!qq .c». L I, 1980) ~ ad fORW 2bn Wuoo7 [ont . () CD ~1n :@ho cegbrd Cado ht*cnez ~asy ikD cdW=c omg 01o n e nr r . WIho s2hr o osoco .f lad daotd .lo~n oro deCom waýh by ik0 k=01 "builht ~ees. Sco essmitcru 2Dr-B-Z~:3M und &CQ0h0Lo DZbthd;. BOugeC-Me (Poderdl Q~idn ~ad), [Eman m . D ata3 en oe~ f ntru~tr mr m"o "eM thoLadon ctna n of &ho L~n o7 North Rh~OirWo lfol. Аппех С Page 11 of 31 т�►в�� в. coмPni:isoи oF иFгвл►и вЕаисеs sтАиО�лоs 1и мОRоссо avo отнея соиитwЕs COMPAitLSON OF STAЛIDARDS АМОКG D1FP'BRBM COU1V?R1SS SOCIAL INFRASTRUCIUR&S MOROCCO TUMSlА FRANCB ITALY CBRMANy 1. ' дакG л�тiг .ооо опо 1.аоо s ооо з т2/inhвb. 2.00 1. 0 1.60 У$СНОО Jm 13 000 1 300 В,230 8,000 000 у ь. 2.и0 1.оо 1.го . т�аI•i scHW� т�1мЬJ� зо.ооо w0 12.Ооо 20 ооо мЬ.л. з.оо 1.гю а.1о Тклп�тдvс cPrrrFтt т�м. rrou го ооо т 3. L(�Ат• ЯЕ ГЕ1`ГтF.R 1дЪвЬ.Аиit 13 000 � 1?.S Э 000 иЬоЪ. .04 0.06 0.20 . Нс�3Т'тТ ц►ЪаЬллiс 45 0 5,000 ?д 000 tn7J{аЬвb. 0.02 1.20 .УО lnhab./ua�t SO 000 1 300 13 700 мц Ь. 0.04 0.06 В. WOMB N1цt lebaehnrt 30 OOf1 м иЬа 0.04 • CU1-7'cЛtAL 1'1iR 1oltisb.lonlt 43 000 1 SOp т2,В00 6,00() 2Q000 п17liаЬвЬ. О.Ов U.Ob O.fO 10. А йhabJwl 10.000 � 000 1 �00 3 000 mha 1i 4. 111të2.IGi0 S CENiER bhиil 20 7 000 3, O.W 0.1! 0.7D 12. С A1tKS lnhab �f 000 3 000 18 000 1.OOU S 000 п�?liabab. 222 0.57 10.00 Э. ''1' inbabJьoц 100,000 10 WO 13.700 1.OOD 3 000 т2linlve. 230 11. 0 0.50 14. C:L� bJ 43,OOU 4 OW 11 . 0 a�nhвb. 0.04 0.08 �ouROe: !ее tвЫе 3 Annex C Page 12 of 31 Current standards of street width, parks and other community facilities seldom allow for effective land utilization ratios higher than 50 %, whereas optimal ratios for low-cost housing projects are around 65%-70%. This explains why even projects specifically oriented to low-cost housing have lower densities than necessary, and consequently provide too few plots per hectare. The average planned densities for low-cost housing projects are rarely higher than 400-500 inhabitants per hectare, and provide 30-40 parcels of land per hectare. A more efficient use of land would require a density of at least 600-700 ( and up to 800) inhabitants/hectare and a minimum of 70-80 lots per hectare. If standards for lot coverage, floor/area ratio, building height and community facilities are also considered, the effective use of land for residential purposes is restricted to a mere 35-40% of the total available surface area. Planning standards, however, are only an overall reference for the design of land development projects. Their effective impact can only be detected in the practical design of the land development scheme. This is why greater flexibility in their application could help promoters in defining schemes which would take into account the financial resources of the beneficiaries. Some recent reforms have in fact begun to consider this issue. Recent urban planning legislation (Dahir N. 1-92-7, art. 21) has built more. flexibility into the implementation of projects, leaving room for schemes with lower infrastructure standards. Although this provision is certainly encouraging, the issue of standards still requires further study and evaluation. The main aspects of standards (see table 6) to be examined are the following: a) the minimum size of land parceLs. The minimum size of parcels prescribed by Moroccan Plans (60 m2) for low-cost housing appears to be appropriate for this type of development. The Team-Maroc study, however, shows that the average size of land for traditional low-cost housing (habitat marocain) tends to be higher (on average 106 m2). This is confirmed by a CIH study, as shown in table 7. Annex C Page 13 of 31 Table 7. Ratios of land use for different housing types (High-rise = Immeubles; Villas; and Low-cost = Habitat marocain) High rise Villas Low-cost Average lot surface m2 300 370 120 % built area 58.3 40.5 81.6 Average built surface m2 830 280 150 Average Floor/area ratio 2.8 0.75 1.25 Average unit size m2 119 200 94 Source: CM, 1992 Moreover, minimum standards for other developments, especially in zones B (Zone urbaine A immeubles allign6s) and C (Zone urbaine d'habitat discontinu) could be reduced in order to achieve higher land use efficiency. b) the lot coverage ratios Lot coverage ratios are an important indicator of the intensity of land use . In the case of traditional low-cost housing (habitat marocain), the Team-Maroc study (Team-Maroc, 1994) calculated this ratio as ranging from 95% to 100%, which is in keeping with this housing typology. The CIH study essentially confirms this information, albeit with a different coverage ratio (82%). As the lot coverage ratio is strictly dependent on the building design, a general land use efficiency principle would suggest that promoters should be freed of this limitation. Similar considerations may be made for B zones, while in C zones , where maximum lot coverage ratios are significantly lower (between 25% and 40%), an increase would be advisable, in order to avoid excess land consumption for this type of housing. c) floor area ratio and maximum height Floor/area ratio (COS-Codfficient d'Occupation du Sol) is generally considered the most important indicator of land use efficiency. In general, floor/area ratios (COS) in Morocco appear to be rather low, especially given the extreme land scarcity. This is certainly the case for low- cost housing zones, where the CIH study reports an effective COS of 1.25, but also applies to high rise buildings, whose average COS is 2.8, and villas, with a COS of only 0.75. Similar considerations may be made for the height standards of buildings. Empirical data collected by Team-Maroc show that over 75% of the total traditional low-cost housing stock, i.e. over 60% of the entire national housing stock, has two floors or less (R+ 1). Conversely, Annex C Page 14 of 31 only 16% of the high rise housing stock ( i.e. 1% of the total national stock) has 6 or more floors (R+5). The extremely low density land use pattern revealed by these figures may well correspond to the preferences of solvable demand, but it is also one of the major factors in current housing deficits and in the extremely high pressure on land. As the empirical data collected match quite significantly the official standards, it must be concluded that a basic change in this policy would be required to achieve better land use efficiency. As a general rule, maximum COS should be increased in low-cost housing areas to at least the value of 2.5, if the promoters of this type of dwellings are to be given greater freedom. High rise buildings need a greater COS, and it is suggested that their promoters should be free of limitations. On the other hand, it must be acknowledged that high rise buildings are not the only solution in designing higher land use efficiency projects. In fact, a high land use ratio (up to a COS of 2.5) can be achieved with traditional Moroccan high-density housing types, as suggested in the recommendations by the Team-Maroc study (Team-Maroc, 1994). d) public services and conununity facilities (6quipements socio-collectifs). Standards for public services reflect the aspiration to satisfy the needs of a well managed urban society. Consequently, they have become an important component in Moroccan urban planning. The official standards, issued by the Ministries concerned (Education, Health, Sport and Youth, etc.), now constitute a basic tool for planners and city managers. Their practical application, however, poses a number of problems. Firstly, standards determine a clear-cut difference between large and small projects. In practice, while large projects need to meet prescriptions fully, and therefore must allocate a significant share of land to services, small projects do not contribute pro-quota to the provision of standards, and therefore have a higher efficiency. In these cases the task of creating the necessary public services is usually left to the public administration: this may explain some of the delays in the implementation of plans mentioned above. Secondly, the automatic application of standards practiced in large development projects increases rigidity and the cost of land and the housing supply system. In most cases, the amount of land prescribed by official standards can be adjusted to the effective needs, and even to the preferences of the beneficiaries as they spread their investment among different urban services or over time. This point is confirmed by planning practice in Morocco. Both SDAUs and Master Plans usually adapt the official regulations to the local situation, as an examination of many local Plans reveals (see, for example, the SDAU for Agadir or the PA for Skhirat). It is surprising therefore to find a flat application of the official standards in many of the low-cost housing projects promoted by ERAC or even ANHI. Annex C Page 15 of 31 e) streets and parks As mentioned above, the use of excessive standards (and especially street width) is among the most important causes of the low land utilization ratio. Currently official standards for streets (20-30 m. for primary streets, 12-15 for secondary, 8-10 for tertiary street) are applied to all types of large-scale land developments in Morocco. As with the other standards, these design prescriptions seem to require adapting to the effective needs and the economic capacity of the beneficiaries. While not a significant loss in areas with low car ownership and circulation, reductions in street width will increase the available parcels of land and therefore reduce their cost. Moreover, evidence of the potential impact of reduced standards on land and housing costs is already available in Morocco. Unauthorized neighborhoods (clandestine) obviously can not be considered a model of land development: they provide very low infrastructure and social service standards and are usually characterized by considerable discrepancies between internal and external hygienic standards. What they do reflect, however, is the spending capacity of a large number of people, and they seem to provide at least a second best choice for a growing number of urban dwellers. The recent Team-Maroc survey highlights some interesting features of this kind of land development, which deserve to be taken into account in reforming the official land delivery process. As table 8 illustrates, while the average illegal development lot size is only slightly smaller than the corresponding legal parcel, its cost is significantly lower (170 Dh/m2 compared to 546 Dh/m2). Needless to say, part of the difference is explained by a lack of infrastructures. Table 8. Comparison of legal low-cost housing and illegal housing. ILLEGAL LEGAL Average lot size 96 106 Cost of land Dh/m2 170 546 Construction cost Dh/m2 680 1,480 Final cost of unit M2 780 1,832 Total cost of housing unit 58,300 193,000 Source: Tcam-Maroc, 1994 As is generally known, illegal settlements have another comparative advantage: much lower construction costs. This is due to piecemeal building, a considerable element of self help and Annex C Page 16 of 31 lower quality finishing. Consequently, an illegal housing unit costs almost a quarter of its legal low-cost equivalent, and is therefore much more attractive to most new city dwellers or newly formed families. The different level of standards in illegal settlements compared to legal housing projects is the key to higher cost efficiency. The adapting of current standards to what may be described as an explicit and solvable demand appears therefore not only necessary, but also feasible. Evidence that this concept is already part of official policy comes from the regularization operations recently launched in many parts of the country. For example, the regularization project for the Oued Eddahab area in Sal6 proposes raising the standards for 43 hectares of existing illegal development by increasing the land used for public facilities from 4% to 11% and lowering the population density from 1,033 to 825 inhabitants/hectare. Table 9: Sal&. Oued Eddahab regularization project BEFORE AFTER Gross land area hectares 43.13 43.13 Urban services hectares 1.740 5.105 % on total land area 4.03 11.83 Net land area hectares 41.29 38.02 Number of lots 3,355 2,687 Lots/hectares 81 70 Population 44,452 35,602 Density/hectares 1,033 825 So rce: Prefecture de Sali 4. The impact of subdivisions on land supply. Although the importance of zoning and of standards cannot be underestimated, what most influences land supply is the actual practice of land subdivision. In other words, it is the practical choice of the residential mix, densities, public spaces and services which determines land use efficiency, as is pointed out by a World Bank paper (World Bank, 1988) . Since a general appraisal of land subdivision practices in Morocco has not been conducted in recent years, a preliminary estimate of the main design orientation of land development promoters may be based on a number of case-studies collected in Rabat-Sal6, Casablanca, and other locations. Annex C Page 17 of 31 This analysis should be considered partial, since it only concerns large-scale public land development projects. Nevertheless, as this type of development represents over 60% of the total ongoing land conversion projects, the sample selected may be considered representative of current trends. The projects selected include the four major public projects underway in Sal6, (Arrahma 1 and 2, Sidi Abdellah, Al Mohit), two large-scale projects in Casablanca (Bachkou and Ain Harrouda), and three smaller scale projects in minor cities (M'Jaraa in Jorf El Melha; Hay El Mohammady in Taroudant; and Al Arz in Khenifra). Tables 10 and 11 summarize the main features of these projects in terms of land use, infrastructure and density choices. The first interesting result of this analysis is the extremely low incidence of saleable land on the total. In all the cases considered, actual saleable land area is less than 60% of the total available land, and averages at around 50%. This depends directly on the high design standards chosen for streets and public spaces, which cannot be justified in low-cost public housing projects. Annex C Page 18 of 31 TABLE 10. LAND USE AND DENSITIES OF HOUSING PROJECTS HOUSDIC PROJEt"~TS IN SALE. ANALYSIS OF' D>EMITIES AND LAND USESE__ RJECT ARRAHMA 1 AltRAMA Å mvDI ABDELLAJK AL MURIT % % % ALLAND AREA A 143.3 100% 103.8 100% 3$9.1 100% 192.7 1004 ENTl. IAND HA 52.5 37% 34.6 33% 101.5 28% 81.4 429 FASTLLAND HA 23.3 17% 26.6 26% 58.5 16% 31.5 16 - -AS--E 2 R ISES HA 3.7 3% 0 0% 11.4 3% 3 35 ADS AND P~ARS 63.3 44% 42.6 41% 187.7 52% 74.8 39 AL POPULATION 73.119 46,379 115 464 69.720 .A tr!.lsATION% 55.8 -9.0 47.7 61.2 MALA.K AREA % 39.3 33.7 29.7 45.3 VROSS DENSITY INHAMA 524 447 322 362 USING PRJECTS IN CAAsLANCA. ANALYffs OF DEMsrMIN AND LAND USEM PROECr OPERATION BACHKOU AMN äARROUDA BACICKOU 1IDI MAAROUF1 F % - - % % AL LAND AREAIIA 22.0 ####o 0 7.0 ### 145 100% SIDENT1IAI..AND1A a. #OUM## £.a. ####### 60.5 42%_ NFRASTL LAND HA nIL, #### n.. ######M 14.12 10%1 pHEmtMUSESIIAn.a. ##### n.a. 0#~100#4 14.9 10% ORVADS AND PARKS c.####### 55.5 38% RIM-ALPOPULATION a.7,100 50 000 ______L__SATION%___ - .61.7 ALEAB15. AREA % as n.a. 52.2 OSS DENSITY INHAB/HA - 1014 345 11ER HOUSNC PIRJECTU I MOROCCO ANALYIS OV DENSITIES AND LAND UslE _ _"rOJ_' M'JAARA MaT Zr.m~sunus: AL ARY _____________JORP EL MELRA TAROUD>ANT EHENIPRA --- %%% TALIAND AREA2A 21.0 100% 38.0 100% 16.8 100% DIDENTIA.IAND H1A 6.6 31% 17.9 47% 8.2 49% FR ASTR. IAND H^ A4.0 19% 1.81 5% 0.1 1% .R_USES ___A_ 0.0 0% 0.0 0% 0.1 1%1 OADS AND PARKS 10.4 50% 18.3 48% 8.4 5 AL POPULATION 6.444 16,826 30,800 UTILSATON% 50.4 51.9 50.0 ALEADL ARFA % 31.6 47.1 4R.8 RON DENSITY INIIA/HA 308 443 643 9r. Pkrdurm d 7 NaI& SNEC, AUC. Annex C Page 19 of 31 TABLE 11. SELECTED HOUSING PROJECTS. TYPE OF HOUSING AND LAND USE HFOUSIG PRo.IC-rS IN SALE. LO DENSITY AND TY1E OF HOUJbNG TINITS ont Afflu~ ARRMLA A2 SMDIADELI.4H ALMORIT i% % l% % TA.PMo.TCTARFA 143.3 103.8 359.1 192.7 of LOTS 6671 3.403 8,21 - s0 r/ 46 33 23 27 .OFiHOUSINCI INITS 13,658 8.433 20.94 1 50 M01 MSING UNIT./HA 95 81. 5e 65 f.SIDENTIAL.I.ANDHA 52.5 100% 34.7 100% 101.6 100% 81.3 100 RELOCATION 45.6 87% 16.5 48% 2.5 2% 19.2 .W-COST IOUNING 0.0 0% 12.0 35% 55.9 55% 13.6 1 IHfi1m RISE HIOUSING 6.9 13% 2.7 8% 19.2 19% 16.5 VIll.Asi10% 24.0 24% 32.0 SING PROJECTM IN CIASA~LANCA lwr DENsITV AND'YPE OF J HN UNT OPERATION BACHKOU AIN pARROUDA SAcOU gIDIMAAROUF ALPROJur AREA 22.0 7.0 145.0 OF LOTS aa, ua. OF HOUSINOUNrnI na. Da. OUSIN0 UNITS/HA• IW.NhrtAL LAND HA 0na. ^M a. U### 60.3 100% RELOCA'11N sa. W## n.a. ###fi 28.8 48% 14W-COSTIIOUMI ra. l0# a. #~Woff 1.6 3 floG RISE IIOUSI4 0 a. ## aa. lM 30.1 50% vILLAN rna. #8## ana. f#IUJ 0% HOUSING PRWECTSIN MOlOCCO LOT DENITY AND TYPE OF HOUSIN UNTS OJE.CT.MAE ar'aaULsMoAMMaDI az r. Mu ATmoamm uaDnrAA 'OTALROJECT ARi,A 2D.95 3 16.63 N.Oy LOTS 517 1352 894 >r5/HA 25 35 55 N.OrF COUS ING Ur# I 0I N TRo.e. -. OUSINO UNrj'/A 54 - 1.. - ... IDENTIAf. lAND HA 6.2 100% 17.9 100% u. 100% RKICATION 0.0 0% 088 5% 0.8 10% LOw-COST HolJSING 3.3 61% 10.87 61% 3.0 62% HION RISr IIOUNING 7 2% .46 31% 2.3 28% VILAS 0.7 11% 0.7 4% 0% o~rce: Preficuam da SW. SNEC, AUM Annex C Page 20 of 31 Similarly, the percentage of land reserved for services and community facilities appears to be high, given the nature of the project. Where public services are included in the project, this percentage ranges from a maximum of 26% of the total land (Arrahma 2) to a minimum of 10% ( Hain Harrouda). A more detailed analysis of the project reveals that this high incidence is directly generated by the flat application of the official national standards, without any adapting to the local context or to the nature of the project. As the combined result of these design choices, residential land is reduced in most cases to less than 40% of the total, the lowest figures being 33% (Arrahma 2) and even 28% (Sidi Abdellah). The resulting gross densities are, as a consequence, rather low, ranging from a maximum of 643 inhabitants/hectare at Khenifra, to a minimum of 322 at Sidi Abdellah. Al Mohit, an ANHI recasement project in Sal6, has only 362 inhabitants/hectare, compared to the 825 inhabitants/hectare for the neighboring Oued Eddahab urban rehabilitation project mentioned above. One of the main reasons underlying the extremely low land use efficiency of these projects lies in the mix of low-cost/low density residential units and of large land consuming building types, such as villas. As shown in table 11, this housing type does not figure in all the projects. But where it does, such as in Sidi Abdellah, it takes up about 24% of the available residential land area, limiting therefore the development of a larger number of low-cost housing lots. In the ANHI recasement project at Al Mohit, villas account for as much as 39 % of the total residential land. The basic argument for the development of this housing mix lies in the mechanism of perequation: larger returns from the sale of high income lots can be used to subsidize low-cost housing or the recasement operations. But this policy not only diverts the public effort from its real target, it is not even coherently applied in the project schemes analyzed. In the Sidi Abdellah subdivision, the price of the land used for villas has been fixed at 1,200 Dh/m2, while the price of land for low-cost housing is 1,500 Dh/m2. In Al Mohit, the corresponding figures are 1,000 Dh/m2 and 1,500 Dh/m2. Thus in Sidi Abdellah, where villas occupy about 24% of the available land, they only pay for 18% of the total costs, whereas in Al Mohit they have 39% of the total land and contribute 30% to the total costs. Although this price structure may be explained in technical terms by the higher density and costs of services needed for more compact housing sectors, the project could be much better balanced financially by converting this high land consuming housing type into smaller low-cost parcels with a higher use density. As a result of these choices, the basic land use efficiency indicators presented in table 11 are extremely low. The number of lots per hectare is under 40 in almost all cases, the lowest case being Sidi Abdellah with 23 lots and a total of 58 housing units per hectare. Annex C Page 21 of 31 Table 12 summarizes the effects of these choices for the four housing projects examined in Sal6. While the average lot size in some of these projects (Arrahma 1 and 2) is close to the average recorded by the recent Team-Maroc study, two other projects (Sidi Abdellah and Al Mohit) present unusually high average sizes for a low-cost housing land development. Finally, the total amount of land required to create one lot of buildable land appears to be excessively high: 367 m2 in Al Mohit, and a record 421 m2 in Sidi Abdellah. Table 12: Average lot size and land needed for the production of one lot PROJECT SURFACE TOTAL N. OF AVERAGE LOT SURFACE HECTARES LOTS SIZE NEEDED PER LOT ARRAHMA 1 143,4 6,671 78 214 ARRAHMA2 103,9 3,403 101 305 SIDI ABDELLAH 359,2 8,521 119 421 AL MOHIT 192,7 5,250 154 367 Source: Prefecture of Sali, 1994. A more detailed evaluation of these projects would require a comparison with the general urban planning provisions to explain the mechanisms underlying these unsatisfactory land use efficiency figures. Such a comparison has been made in the case of Sal6. The results of the analysis are summarized in table 13, which shows the planning prescription for the three main development sectors of the city: Sidi Moussa S-0, Quartier Industriel and especially Sidi-Moussa N-E, where all four development projects examined here are located. The data clearly show that the maximum allowed densities in different planning zones are significantly higher than the densities resulting from the project choices of the four developments. In particular, the maximum allowed density of the urban area (592 inhab/hectare) is over 50% higher than the average density of the four projects combined (382 inhab/hectare). In fact, the four low-cost housing projects present the lowest densities of the entire urban area of Sal. This confirms that urban planning plays a lesser role in restraining land use efficiency than actual land development design choices. Annex C Page 22 of 31 TABLE 0. SALE. COMPARISON OF LAND USE PLAN AND aUeSDiso DEN29TE SARE.FcMPARON 3ETWEE0 DENgaTrE A LLOWZD IBYTE R PHEM SEBH-McUSSA SUED=SDows S1= gZONiNG lSURFACE ha IHOUSING UNITS A./ha SIDA-MOUSSA S- B2 3.7 146 775 17-15-13 B3 1 5.7 1221 6461 B4 |0.9 98| 517 C2 10.7 ___127 672 E _ 1.91 78 4231 22 34.4 18 ¯ 9 TOTAL _7.3_23377717 TUSTIÌAL SECTOR _ ¡ 133 11.3 126 666 C2 i.2501 7951 ITOTAL 21470.2]6|70 ISDEMOUSSAN-ST ¯ 2-5_2 _ 4.6 26 7 B3 |26.7 1221 _6 c2 J 12.6 117| 6201 Di 29.4 16 03 TOTAL J_112 592 marna | | £43.4ls - sw2 2 _ n@.9f1 -li_j 4 läfiAEDElLA | 5.2 S0 3221 a4 _____ l19l Sourc=: Préfcu of Salé SCalculaed on net re al land ama Annex C Page 23 of 31 5. An exercise in land use efficiency International experience shows that land costs can be reduced in low-cost housing projects by improving the design efficiency and by adjusting the choice of the critical variables to the priorities and economic capabilities of the project beneficiaries (World Bank, 1988, and World Bank, 1989). With respect to these goals, Morocco has a long tradition of low-cost housing schemes, which promoters and project designers must reconsider as an experiment in the practical implementation of projects. What follows is a numerical exercise proving that a much higher cost and use efficiency of land can be achieved in housing projects by adjusting the design variables at stake. The exercise is based on the application of a model (Wbrld Bank, 1988) which makes it possible to evaluate the effects of variations in the design choices of a given project. The application of the model to two of the four low-cost housing projects of Sal6 already examined (Sidi Abdellah and Al Mohit) enables us to check briefly possible land use efficiency strategies. Tables 14 a-b-c-d illustrate the effects of the application of four different strategies. To illustrate more clearer the potential of each option, they have been applied individually. The four strategies are: a) reduction of the land area used for circulation and parking (table 14a) The reduction of land used for circulation, parking and other public non-commercial uses (parks, etc.) yields the greatest increase in effective land utilization. In the two cases examined, the number of parcels and the density of population sharply increase as land reserved for circulation is reduced. In the case of the Sidi Abdellah project, this increase is of the order of 100% as land reserved for circulation is cut from the current 52% to 25%. Correspondingly, the net cost incidence per m2 of land decreases from 1,238 Dh/m2 to 749 Dh/m2 (a reduction of about 40%). In limiting circulation space, attention must obviously be paid in the design phase to maintaining the necessary spaces for access, future changes in circulation pattern and hygienic standards. One possible way of compensating for the negative effects of reducing public space is to open up in a regulated way public spaces pertaining to schools or other urban services. Annex C Page 24 of 31 b) Reduction of the land area reserved for the villa housing type (table 14b). As mentioned above, this housing type does not fit in with public intervention targets in the land and housing markets, nor is it economically advantageous in the framework of low-cost housing projects. The alternative examined in table 14b shows how the straightforward elimination of this land development type increases low-cost land parcels by over 22% in Sidi Abdellah and by 45 % in Al Mohit. c) Reduction of land area for urban services (table 14c) Official urban services standards are often applied without any attempts to adapt to the local context. These standards could be adjusted, therefore, to lower levels of land consumption, in order to reserve a larger area for residelitial needs. In the cases examined, the reduction of the land area for urban services from the level originally planned down to 15% of the total built area produces an increase of over 50% of the number of parcels and density in Sidi Abdellah, and a corresponding increase of 38.5% in Al Mohit. Land unit costs consequently decrease. As in the case of circulation space, the decrease of the land area used for services can be mitigated by design criteria making optimal use of the spaces needed by the various services. d) Increase of land use density (table 14d) Land use density can be increased by allowing higher buildings with more housing units, or by increasing the land utilization ratios to facilitate the diffusion of high density traditional constructions. The test conducted shows a clear decrease in the land incidence factor, when the number of allowed housing units is set at 5 per lot. This exercise should be considered as a theoretical test, since in actual practice the design variables are evaluated and combined at the same time. Nevertheless, to illustrate the potential land efficiency improvement of this process, a combination of the strategies suggested here was tested in two of the projects. In particular, circulation land area was cut to 35 % of the total area, the land area for villas was eliminated, public services land was set at 12% of the total, and the average density of housing units per lot was left at 2 in one case, and increased to 2.5 in the other case (see Tables 15 a-b and 16 a-b). These variations have been done only with a demonstrative intent, and would obviously require an overall impact analysis based on the actual physical and infrastructural setting. The exercise nevertheless shows that, although these design solutions do not correspond to the maximum land use potential, the effects on project efficiency are quite pronounced. Density in Annex C Page 25 of 31 Sidi Abdellah reaches 606 inhabitants per hectare, compared to 316 in the original project, and net land costs are down 29% to 876 Dh/m2 compared to the original 1,238 Dh/m2. In the case of Al Mohit, density rises from 362 to 767 inhab/hectare, and cost is down 13% to 985 Dh/m2 as opposed to the original 1,131 Dh/m2. Annex C Page 26 of 31 TA. 16a. STRATEGIS FOR THE REDUCTUNG OF LAND COST 9NCDENCE. REDUCTION OF CURCULÄTBON AREA [AREA|| INHAB/HA H/ RODB ABDELLAH__ _ _ _______I____ 52.2 3 ,521 318 1,238 -----_oT_ _-~--0 397 1,056 M2 40.0 12,196¯ ¯[ 3 958 ff3l 35¯.¯ 13,7¯~¯¯ 60 M - -3f0.0 15,208 j 564 808 -5 25.0 16,713 || 620 749 C 38.3 ]j 5,250 [ 325 ,3 Hi 1~~- j[ 35.0 5.657 390 1,0e H2 30.0 ,272_432_95 source: MfeeTeur- ov slr TA 0 . STRATEG8EM FOR TWE REDUCTOOM OF LAND COST 8WC8DENCE. REDUCTUOM OF LAND FOR THE "MALLA"9 MOUSMNG TYPE LAND FOR N. LOTS LOT/HA |DENSIT VILLAS INHABIHA IDI ASDELLAM ROJECT t 23.50 8,521 23 316 1 15.0o 9,149 25 339 F_2_ ~~T- 1.o0 jp 0,519 27 353 3 \\ 5.00 \ 9,947 28 369 4 0.00 10,414 29 3e8 oL R9OHDT #ROJECT 38.80 5,250 27 362 1 30.00 5,650 29 389 2 1 20.00 6,188 ~1 32 _F 426 3 0.0 6,028 35 11 471 H4 0.00 7,616 ¯¯40 525 Sourco: Préectur l of Salé Annex C Page 27 of 31 TAB 14 C. STRATEGIES FOR THE REDUCTION OF LAND COST INCIDENCE. REDUCTION OF STANDARDS CIRCULATION'LAND FOR NW. LOTS DENSITY NET COST AREA INFRASTRUCT. INHABIHA DHM2 % ON TOTAL I BUILT LAND IDI ABDELLAH ROJECT 52.20 34.15 8,522 316 1,238 1 52.20 30.00 9,123 339 1,056 52.20 25.00 9,842 365 958 3 52.20 20.00 10,552 392 878 4 52.20 15.00 12,810 419 808 HS 52.20 0 13,439 499 748 MOMIT ROJECT 38.32 26.50 5,250 382 1,131 1 38.32 20.00 5,746 396 1,068 2 38.32 15.00 6,126 422 985 3 38.32 0.00 7,261 500 914 Source: Prffecture of Salb TAB. 14 D. STRATEGIES FOR THE REDUCTION OF LAMD COST INCIDENCE. INCREASE OF LAND USE DENSITY DENSITY N. LOTS N. HOUSING DENSITY NET COST LAND COST HOUSINGILOI UNITS INMABMA DH/M2 PER HOUSING % UNIT IDIABDELLAN ROJECT 2.46 8,520 20,959 321 1,238 60.253 1 3.00 8,520 25,580 391 1,238 49.404 4.00 8 34,080 622 1,238 37,053 3 5.00 8,520 42,600 652 1.238 29.642 6.00 8.520 51.120 783 1,238 24.702 MOHIT ROJECT 2.46 5,250 12915 362 1131 89.324 1 3.00 5,250 15.750 449 1,131 73246 H2 4.00 5,250 21000 599 1.131 54,934 H3 6.00 5,250 26,250 749 43,947 _4 6.00 5,250 31,500 800 1 131 36,423 Source: Pr6fecture of Said Annex C Page 28 of 31 Ta5iJE i9 A. AALYSIS OF S0D A!IDLLAN4 SU'CISIONI IN S E.POJECY AS DESIGNED S4D4 ABDELLAH IN SA COST OF LAN AND fc--Porantoq-.- COST PMIFRASTRUCVURE (V.R.D.) cot Conti- Std:s Fnanci RECOVERY ftn2 n wmm r ooot .U n2 GrooIand GOetfrn2 60.00 35.00 33.00 27.00 19s.B2 Lcnd proportion - 0.00 0.00 0.00 0.00 0.00 In oo Enkcrructxoo 220. 12.00 0.00 27.00 313.69 ORo nofrectuns:s • - 38.20 12.00 0.00 27.00 54.34 Ohr doioloprnt eots - "' 9.0 0.00 0.00 27.00 12.18 MRUCTURED LAND. GROSS COS T70 RECOVER PER M2 017.0200982 LAmD UTIL85ATION AND NOM RWDE7AL LAND k AES PERCENT. 7. N@n-Ra3d, Lod Toal project oroo Ono 200.93 100 ~ SoU pxi infrtm. tnd Co4t~ % 52.20 52.20 % parnot m2 PmopubL Infrtruct% 0.04 9.04 u - Prty o l m2 115,000 3.20 U 500 Seeondry schooo. m2 114,000 .17 S 400 Othnrinfruc^us m2 118,100 3.20 % 42 Comm~rco 01 n2 12,=00 0.33 % 2,500 Cewrc 02 m2 10,000 0.29 u 2,00 Comn= 03 m2 0 0.00% 0 Skndutm2 114,000 3.17% 600 SALEAGLE LAND m2 1,4n9,837 41.7 % RESMDENTIAL LAND m2 1,016,337 02 q$ NUMBER OF LOTS 5,522 Avarog PeroRat: 13 LOTS/HECTARE 24 TAL POPULAION 113,579 POPULAT.DEs:TSmRo 890 NFRTRUCTURED LAND. RET COSTS TO RECO~R PER Ma 1289.10 äcuo: owboraton of deta cupp~3:d by th3 Profcctuo o So TAOLE IG. ~VALYSIS OF 21m O ASNLLAN ON SAL. PROJEcT MITH CANGES S IN SALE CVT UA AN Bo cPC@clg:l-1 COST WEIRaGTRutvURmm (V.i.D.) coo Ca:dbb- snb FenrocMo RECOVERY ftn2 ganceso a ethor oooto . 5san2 bnd ou/m2 60.00 S3.00 33.00 ZT.00 13.82 Lord pcporton 0.00 0.00 0.00 0.00 0.00 in vR aructro 220.33 12.00 0.00 - 27.00 313.69 Off utto Infrtnooire - wm.20 12.00 0.00 27.00 54.34 Otc=rv orc:ss D.C0 0.00 0.00 27.00 12.19 NFR1ATRUCTURED LAND. GROSS COSTS TO RECOVER PER 42 017.03 LAND UTISATCMN AND NON REUENTIAL LAND SALES - PERCEMn T. WN~c29d. Land Tand pojec c7 ha M.36 100 %5 Sol pos ifrotr. t2m Ctreslan % 35.00 3.00 % por no9 M2 Porhoopubt infstruc% 6.04 4.80 % P?~ri echen! m2 115000 2.50 rs 500 Sc~ndory o3choof. m2 114.000 2.60 t 400 frcotr m2 118,500 2.0 423 ~m 01 m2 12,000 0.20 u 250 02 m2 10,000 0.20 % . 2,500 Conm~CO 0m2 0 0.00 u 0 Smell fnduo . m2 114,000 3.17 qS 800 SALEABLE LAND m2 2,117,07 69.%9 % RESIDENTIAL LAND m2 1,634,107 46.20 u NUMBER OF LOTS 19.745 Av~rce Po~ta±: 13 OTS/ECTARE 47 OTAL POPULATION 217,92 POPu!ATDNSITYRa?o 000 bIwmR4TLUCTURED LAND. NET @OmTn TO R2COV R PeR 2 07.01 S.0reo: Omburtio of a2to OuppScd by~ 01h Prefechao al Sel Noof Ch~i bn tha pro): cm tho fento*a: o) dmi:tlan orco S5; b) minten~ Of lit c%M or N8Ico c)b f änodsr 12 oftocoý&2£Caege; d) meam93 o 2 F2n ptt Annex C Page 29 of 31 TASLE 18 A ANALYSIS OF AL MOHrr SUBDIVISION IN SALE. PROJECT AS DESIGNED AL MOJIT IN KALK COST OF LAND AND as. I-P.rentage-| COST INPRASTRUTURES WMJD.) cat. Contin- Studis Fnancial RECOVERY 4n2 gmneff 8 ofer ot~ . s~m2 Gros land cot/m2 100.00 35.00 33.00 32.40 237.72 Land pMphration - - 0.00 0.00 0.00 0.00 0.00 in skl Infrastrcturs 221.41 12.00 0.00 3240 328.32 Off sie Inrm tructura " 30.20 12.00 0.00 32.40 56.05 thr developmtrdots * 10.69 0.00 0.00 32.40 14.16 INFRASTRUCTURED LAND. GROSS COOTS TO RECOVER PER MI 636.8 ..===. m..=~ ..... ..... .-----.= ====== ===== ======== LAND UTLISATION AND NON RESIDENTIAL LAND SALES PERCENT. T. Non-Resid. Land Total proect wr ha 393.70 100.00 % Sal. price Infra~r. lund Ciroulatli % 38.32 38.32 % per net m2 Paomk+publ. Infrstruct% 6.40 64 -- Pimary school m2 65,000 3.37 % 500 Secondary swhools. m2 63.000 3.27 % 400 Other infrastructusm m2 64.500 3.35 % 431 Comvere 61 "2 14.000 0.73% 2500 Cowmere 02 m2 5,000 0.28 % 2500 Comfmerce 63 m2 0 0.00 % 0 Sma indust. m2 50,000 2.59 % 00 SALEABLE LAND m2 1,084,574 66.28 % RESIDENTAL LAND m2 823,074 42.71 % NUMRER OF LOTS 1,240 Average PemM/lot: 13 LOT~/ECTARE 27 OTAL POPULATION 69,713 POPULAT.DENSITYiha 362 INFRASTRUCTURED LAND. NET COSTS TO RECOVER PER M2 1131.62 Souroe: elaboration of data supped by the Prefecture of Sai TABE 169 . ANALYSIG OF AL MOIMT SUDDIVISION IN 9ALL PROJECT WITH CHANGES AL MOlItT IN SA1I COST OF LAND AND Base |-Percentages-I COST INPRASTRUCTURER (VJLD.) octa COntin. Studics PmnlI RECOVERY IM2 gmncim &oher oost .tm2 Grss land coat~m2 100.00 35.00 33.00 32.40 237.72 Land prepartion - - 0.00 0.00 0.00 0.00 0.00 In Wto Infrastruotur e 221.41 12.00 0.00 32.O 328.32 Off sita Infrastruturema 3.20 12.00 0.00 32.40 56.65 Other devloprnnt ostsl - 10.69 0.00 0.00 3240 14.16 INFRASTRUCTURED LAND.GROSS COSTS TO RECOVER PER M2 636.85 LAND UTILSATION AND NON RESIDENTIAL LAND SALES PERCENT. T. Non-R~ud. Land Total project arma ha 192.70 100.00 % Sale prinfrastr. land ulation % 30.00 30.00 % per nut m2 Average Parstlot: .40 3.50 %- Prtmury school m2 65,000 2.50 % 500 Secondary schools. m2 63.000 2.50 % 400 Other infrastrutes m2 64,500 2.s0 % 431 Commerce #m m2 14.000 0.73 % 2500 Com0eroe 2 m2 5.000 0.28 % 2500 Comme0rce 3 M2 0 0.00% 0 Smad nust. R2 50,000 2.59 600 SALEASLE LAND v2 1,244,00 64.60 % RESIDENTIAL LAND m2 983.400 51.03 % NUMBER OF LOTS 8,191 Average Perlot: 16 OTS/HECTARE 47 OTAL POPULATION 147.UI T.DENSITY/ha 767 INFRASTRUCTUNED LAND. NET CIMTS TO RECOVER PER MI sas.s SouM: elabmr of dat suppaed by th Prfectre of Sam Mhmng~s kn the project r the fblong a) ~utation arma 30%; b) eIMnsU~n of the arna for 'ias o)lund for oca kIramsuue 12 % cf tta ara; d) .vra desy 2,5 houng unit par lt Annex C Page 30 of 31 REFERENCES ANHI (1993): Etude de faisabilité du projet de Lotissement AI Mohit dans la Municipalité de Laayayda. Délégation Préfectorale de l'Habitat de Salé, TUM Sarl, Novembre. Bahi, Hassan; Alami, Mohammed Hamdouni (1992) Urbanisation et Gestion Urbaine au Maroc, Rabat. Bundesministerium für Raumordnung, Bauwesen und Städtebau, Deutchland (1993) Baugesetzbuch (Federal Building Code), Bonn. CIE (1992) Étude du secteur de la construction, INAU, TUM Maroc, Rabat. Commune de Skhirat (1994) Plan d'Aménagement, Rapport justificatif, Ministère de l'Interieur, Wilaya de Rabat-salé, Préfecture de Temara. Falco, Luigi (1980) Gli standards urbanistici, Roma, Edizioni delle Autonomie. Keene, John: Laghouat, Mohammed (1985) A Comparative Analysis of the Potential Contribution of the Schéma Directeurs of Morocco's Major Cities to the Formulation of the USAID Strategy for Urban Development Assistance, prepared for USAID, University of Pennsylvania, Philadelphia. Lahbil Tagemouati,Naima (1994) Le foncier: autopsie du prix et de l'échange. Le cas de la ville de Fès, Wallada, Casablanca. Lehzam, Abdellah (1994), Le poids du foncier dans le cout de construction des logements urbains: le cas de quelques centres urbains, Rabat, INAU, Les Cahier du CERAU. Ministère de l'Habitat (1992) L'habitat en chiffres, Rabat, Direction de la planification et de la Programmation. PADCO (1987) Etude de la politique foncière urbaine au Maroc, preparé pour le compte de USAID, Washington D.C. Préfecture de Salé, Délégation de l'Habitat (1994) Fiches et montages financiers des opérations, Salé. Rivkin Associates (1985) Urban Development Assessment, prepared for USAID, Wasington D.C. Royame du Maroc (1992): Proposition pour l'habitat du plus grand nombre. Synthèse, Rabat, Cabinet Pinseau, Janvier. S.A.E.M. (1992) Schéma Directeur d'Aménagement Urbain de Taza, Ministère de l'Interieur, Direction Générale de l'Urbanisme, de l'Aménagement du Territoire et de l'Environment. Team-Maroc (1994) Étude sur les coûts de production du logement, (6 Fascicules) , Ministère de l'Habitat, Direction Générale de la Planification et de la Programmation, Rabat. World Bank (1986) A Model for the Preparation of Physical Development Alternatives for Urban Settlement Projects (The Bertaud Model), PADCO, EDI, Washington D.C. World Bank (1986) Gestion Foncière au Maroc et en Tunisie, Actes du Colloque, Barcelone, 22- 25 Septembre 1986 World Bank (1988) Efficiency in Land Use and Infrastructure Design. An Application of the Bertaud Model, by A. Bertaud, M.A. Bertaud, J.O. Wright, Policy , Planning and Research Staff Discussion Paper. World Bank, (1989) Malaysia: The Housing Sector. Getting the incentives Right, Asia Regional Office, Washington D.C. Annex C Page 31 of 31 LEGISLATION Dahir n. 1-92-7 du 17 hija 1412 (17 Juin 1992) portant promulgation de la loi n. 25-90 relative aux lotissements, groupes d'habitation et morcellements. Dahir n. 1-92-31 du 15 hija 1412 (17 juin 1992) portant jromulgation de la Loi n. 12-90 relative à l'Urbanisme. Decret n. 2-92-833 du 25 rebia Il 1414 (12 octobre 1993) pris pour l'application de la loi n. 25- 90 relative aux lotissements, groupes d'habitation et morcellements. Decret n. 2-92-832 du 27 rebia II 14141 (14 octobre 1993) pris pour l'application de la Loi n. 12-90 relative à l'Urbanisme. c:\j\mor\rcponi\houstrat\aniex-d Annex D Page 1 of 12 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Land Price Trends and Instruments of Land Management This report analyses the land issue in Morocco and proposes several instruments to incorporate land to urban use in a manner consistent with free markets. 1. Land price trends. Land as a factor of construction costs 1.1 It is not easy to obtain statistically reliable data on urban land prices in all of the population centers, particularly.in terms of a breakdown of price levels according to the zoning plan or availability of public infrastructure. We have attached a few tables most of which are from the construction survey carried out by CIH in October 1991. The table on land prices and price per constructed m2 was drawn up by Agence Urbaine de Casablanca in May 1994. Other data are taken from monographs cited in the annex, and others, of a more qualitative nature, were obtained through interpersonal contacts. 1.2 The main difficulty is in comparing prices among the three subsectors: legal public production, which most often draws on land reserves of which the purchase cost is zero or very low, legal private production, subject to market conditions, and the production of unauthorized developments, an area in which data can obviously only be obtained by indirect means. An in-depth analysis of comparative production costs in the three subsectors has yet to be performed. 1.3 The level of land prices in all categories appears to be high in Morocco. It is high both in absolute terms and in terms of purchasing power, and seems to have been rising rapidly over the past few years, following a period during which it remained more or less stagnant, even experiencing a slump in the mid-eighties. The following data were obtained during the CIH survey in 1991 (concerning prices of operations authorized in 1989) and indicate price trends by mode of purchase and construction type (these criteria partially overlap). Annex D Page 2 of 12 Trends in average price per m2 by year and mode of purchase (in DH) 1982 1984 1986 1989 Rate of Increase in Annual Prices 1982/1989 Government transfer 189 197 303 238 3,3% Purchase from government 242 516 962 935 21,0% transferee Direct purchase from private 510 849 1028 1205 13,0% individual Purchase through middleman from 600 1159 1506 1368 12,5% private individual Direct purchase from developer 312 648 788 716 11,3% Purchase through middleman from 600 420 538 1018 7,8% developer Overall 328 390 605 815 13,9% Source: Quantitative survey - 1991 It is not surprising that land transferred by the Government should have the lowest growth rate, but what is most noticeable is the very strong growth (21.0% per annum) for land purchased from government transferees, reflecting their tendency to engage in speculation. Thus, the initial transferee is the beneficiary of the differential between market price and transfer price. This result suggests that the mode of establishing the transfer price should be amended and that the establishing criteria should be explained. This impairment could be corrected through the creation of a "Right of Preemption". Annex D Page 3 of 12 Trends in average price per m2 by construction type (in DH/m2) 1982 1984 1986 1989 Rate of annual increase in prices 1982/1989 Apartment buildings 419 783 914 1596 21,0% Villas 226 489 507 696 17,4% Moroccan housing 403 295 513 611 6,1% Industry/Commerce - - 61 381 84,1% Overall 328 390 605 815 13,9% Source: Quantitative survey - 1991 1.4 A combination of the two criteria (mode of purchase and type of construction) confirms the first analysis, but also underscores the extent of the price differential, both for apartment buildings and for Moroccan-type dwellings, between the prices of land transferred directly by the Government and those of lots purchased from a government transferee. Average price per m2 of land by mode of purchase and type of structure (in DH) Mode of land purchase Apartment Moroccan buildings Villas Housing Direct Government transfer 447 288 238 Purchase from Government transferee 1291 452 724 Direct purchase from private individual 1712 670 530 Purchase through middleman from private 1961 833 803 individual Purchase from private developer 912 834 509 Purchase through middleman from developer 1173 777 526 Source: Quantitative survey - 1991 The other comparisons should be analyzed with some caution since they do not include any indication of their -location. It should also be remembered that these prices are underestimated, since the operations analyzed (construction projects authorized in 1989) were carried out on tracts purchased on average 2-3 years before the actual start of the project. Annex D Page 4 of 12 1.5 The example of Agadir, studied by INAU', offers the particular case of the de facto public monopoly (municipality, CGI, ERAC) of subdivision production currently existing in three areas (Dakhla, Najah, El-Massira). Land transfer prices vary between DH 300 and DH 1,200 per m2, and their growth has been particularly rapid since 1980, fluctuating between 15% and 20%. Land prices (DH/m2) per dwelling unit Dakhla 500 - 1200 El Massira 300 - 600 Najab 500 - 750 Such trends in land prices reflect a substantial weakening of land purchasing power, since inflation is presently contained at 5-6% per annum. 1.6 The share of the land cost in the final price of housing has steadily risen over the past fifteen years, according to all sources analyzed, for all subsectors of production. Trends in components of cost per m2 of floor area (in constant DH) 1972-80 1980-84 1984-90 1972-90 Land cost 9,7% -23,8% 7,5% 3,6% Major cities Construction 4,9% -4,3% 2,1% 1,9% cost Average cost 6,5% -7,5% 3,9% 2,5% per ml of floor area Land cost 6,8% -20,0% 9,7% 6,2% Small centers Construction 4,95 -4,3% 2,1% 1,9% cost Average cost 6,5% -6,5% 3,3% 2,5% per in' of floor area In the case of legal housing production, the Team-Maroc study puts the land cost factor at 25-31 %. INAU, Les marchsfonciers et immobiliers & Agadir. Les Cahiers du CERAU, February 1993. Annex D Page 5 of 12 Price of Dwelling Construction (DH/m) Land Cost and other costs Moroccan housing 1628 25% 75% Apartment buildings 2340 26% 74% Villas 2730 31% 69% A more precise study carried out in 27 urban centers will make it possible to refine these overall results, without actually revising them. Particular focus will be placed on land cost levels of, or even exceeding 40% (Casa, Sidi Bernoussi). - 1.7 There is little information on land prices for illegal developments. From various conversations it appeared that those prices could range between DH 100 (poorly serviced peripheral areas) and DH 1,000. Estimates of the land cost factor (varying between 15% and 20%) are only very approximate given the gradual pace of development. On the other hand, it seems that market is suffering not only from the effect of price increases resulting from restrictive government regulations but also from their secondary effects on the different land submarkets, initially triggered by the narrowing (to what extent it is not yet known) of the pool of public land reserves. 2. An operational tool - the Zone d'Am6nagement Concert6 (ZAC 2.1 The largest part of the legally regulated production of serviced land is carried out by the public operators (SNEC, ERAC, ANHI, etc.) which produce fully serviced sites, often with high levels of infrastructure. On the other hand, a large share of illegal private production is often very deficient in infrastructure. The idea of public-private partnership, which has been in circulation for several years in Morocco, aims above all to fill this void by putting in place an operational urban development tool based on a true partnership between the public authority (municipality) and the private operator, and defining the rules of the game in both physical and financial terms. 2.2 Creation of the ZACs is also intended to be a tool for implementing land use plans. One of the present difficulties of land mobilization in Morocco derives from the fact that these plans have no phasing and do not include any operational resources for their implementation. As a result, a public developer may take the initiative to urbanize a site that is in one way or another a public land reserve (and those reserves are being rapidly depleted), or else private initiatives may move in haphazardly, with no reference to the land-use plan. The ZAC aims to fill that gap. 2.3 Also, within this perspective is the idea of a certain degree of government withdrawal from real estate operations and increased intervention by the local authorities, with. the Government exercising more of a supervisory and incentive role. This idea ipso facto involves casting a heaving burden on the local authority/private developer partnership and brings Annex D Page 6 of 12 us back both to the problem of local finances and of infrastructure financing by the private partner. 2.4 The text prepared ("draft law relative to zones d'am6nagement concert6"), an essential component of this reform, comprises two types of operational mechanism: - The development of a zone with a controlled land ownership by a public or private developer within the framework of a partnership with the local authority. This mechanism is similar to the French ZAC mechanism; - Development by the land owners themselves. This idea goes back to the old 1917 legislation on owner associations, and aims to reinvigorate it by introducing a rule of qualified dual majority (in terms of land area and of owners). 2.5 These two concepts are innovative and very important, and should be developed. But they reflect two different development logics and different land tenure situations and land pattern situations: the first concept relates, on the basis of a clear land ownership situation, to a negotiation between partners of comparable influence and technical levels, allowing for the adequate distribution of infrastructure costs (possibly including cross-subsidies). On the other hand, the second idea, development at the initiative of a group of private owners, above all reflects the concern of overcoming difficulties arising out of a complicated situation, and the resistance of small land owners, by using tax incentives and coercion (possibility of overcoming possible resistance from 25% of owners through expropriation). Owners can withdraw from the negotiation in case of non-agreement (inverse condemnation). The merging of these two concepts in the ZAC mechanism makes this a very complex, juridically risky and financially fragile arrangement. They should perhaps be separated in the draft law. 2.6 On the other hand, it is very important to systematically mesh the operational mechanism and the zoning plan. If the latter is to become a true building site production mechanism, it should, right from the time of its approval, contain operational programming projections for at least the next five years. A preliminary land study accompanying the zoning plan would then make it possible to determine which of the development plans will have to be implemented (ZAC, owner association, etc.). But in one case as in the other, implementation of this programming imposes the need for incentives or coercive measures (in France, a ZAC decree opens up the possibility of recourse to expropriation). 2.7 With these two mechanisms covering the problems of land replotting and infrastructure financing, it is possible to envisage supporting them through a mechanism that would transfer land tracts free of charge at the time the operation is set up, designed as a counterpart to the type of construction assigned to the tract concerned. Land involved in this type of free transfer (10% in France, up to 30% in Germany) may be allocated to certain uses, in particular public infrastructure. It may also be developed by the public authority or reassigned at market value to a private developer. Annex D Page 7 of 12 2.8 Development of this operational tool has a dual purpose: to mobilize the supply of land, and to guarantee financing and construction of the principal infrastructure items. It is essential to avoid letting legal complexity and/or the sophistication of the operational tool end up making it unattractive, nullifying the content of the land-use plans and leading to a greater proliferation of illegal developments. 2.9 The draft law could incorporate the two chapters: "Urban Planning" and "Operational Planning", and be articulated in the following way: (a) Land Use Plan The draft law should include (i) the delimitation of zones to be developed or replotted with a timetable, and (ii) a set of rules for these zones with the choice of the operational planning tool, a possible Right of Preemption, determination of reference prices, replotting perimeters, and financing mode for infrastructure. (b) Comprehensive Development Zone (ZAC) The presentation of ZAC should list the following: (i) a definition by the authority of an earmarked zone where expropriation will be feasible; (ii) a brief study of land issues, (iii) identification of developers that will service land and possibly build; (iv) creation of an inverse condemnation (see para 2.5) for owners; (v) illustrated zone development scheme; (vi) free transfer of land easements (in the order of 10 to 30 percent) for the construction of infrastructure; (vii) negotiation of equipment cost share between the developer and the authority; and (viii) definition of the implementation timetable. (c) Replotting Zone The draft law should list the eight following components: (i) definition by the authority of a replotting zone; (ii) estimation of land value, which will guide the redistribution; (iii) commitment of 2/3 of owners, controlling 2/3 of the land; (iv) creation of a corporate entity (which remains to be identified), representing the owners, in order to study and implement infrastructure and recover costs; (v) free transfer of land to the local authority, in a ratio which remains to be determined; (vi) establishment of replotting scheme for urban planning; (vii) recovery of equipment costs from owners, in proportion to their interest; (viii) redistribution of land to owners, related to the initial amount; (ix) termination of the entity at the end of the process. 3. Right of Preemption [droit de priemption] 3.1 The dwindling stock of public land reserves, although perhaps less dramatic than it is often reported to be, obviously constitutes an important factor in the present land supply crisis. It generally gives rise (see for example the Team-Maroc study on housing production and development in Morocco) to a recommendation that top priority be assigned to replenishing the land reserves. Annex D Page 8 of 12 3.2 Local Authorities have different legal means to acquire land. Purchases by friendly arrangement are encountering more and more stumbling blocks, either through an increase in the offering price, or through reluctance to sell (the "wait and see" attitude of both public and private land owners is very evident). 3.3 As for expropriation, while this is clearly defined in the legislation, it involves a very long and complex procedure and is often strongly resisted, and there are more and more instances in which the justifiability of such operations is being contested in the courts. Thus the RS (strategic reserve) zones often exist only in the plans, there being no concrete means of land resource mobilization in this case. 3.4 Mid-way between friendly purchase and expropriation, the Right of Preemption could open up an interesting route for public intervention in the land markets. It consists of giving a public developer, e.g. an urban agency, the first chance, in certain areas, to purchase a plot as soon as the owner announces his intention to sell. The Right of Preemption will allow its beneficiary to acquire land in place of the initial buyer. Widely used in some countries like France and Germany, the Right of Preemption needs technical conditionalities in order to make transfers more transparent and be an incentive and a general motivation. 3.5 Under the 1993 legislation creating urban agencies in Morocco, provision was made for those agencies to have such Right of Preemption, subject to official authorization in a subsequent law. The present draft law embodies an extensive approach to this Right of Preemption, which is explicitly designed to counter speculation. This right may be created in areas slated for restructuring or renewal, in new urban development zones, zones earmarked as land reserves, and areas of large-scale infrastructure. On the other hand, nothing is said about the criteria for use of the assets purchased through exercise of that right. It seems necessary to impose restrictions, based on a public interest criterion (for example resettlement or low-income housing) concerning the permissible uses of land tracts acquired through exercise of the Right of Preemption. 3.6 The draft law provides for a price adjustment mechanism to be used when the beneficiary finds that the announced price is too high. This point raises a sensitive problem and could create serious difficulties. If the zone in question is the subject of tension on the land market, the best strategy is, as soon as the zone is created, to announce a "reference price," on the order of the current market price, but to which purchases already effected under the Right of Preemption would subsequently remain tied. It is important to avoid systematic disputes over prices and the time-consuming procedures likely to ensue. Determination of the reference price should be based on a proper land price observation system. The general principle should remain buying the land at market price. Ties to "reference prices" should limit retention behavior of owners waiting for infrastructure services. 3.7 Management of the Right of Preemption is a sensitive matter, since it means that every proposed real-estate transaction has to be reported to the holders of that right. This Annex D Page 9 of 12 involves large-scale publicity effort at the time of creation of the zones, directed in particular at the notaries, adouls (notaries operating in the traditional system), and land registrars, who should in any case never agree to record deeds concerning transactions that have not been reported to those holding a Right of Preemption. 3.8 The effective use of this tool is linked to the quality of management of the beneficiaries, which are most likely the urban agencies. It is essential to avoid allowing the price adjustment mechanism, which can cause the whole process to end up in court, to make operation of the land markets even more cumbersome. It seems that zones earmarked in the zoning plans for Right of Preemption purchase should be created in small numbers, to encourage would-be purchasers holding that right to come forward of their own initiative. The existence of a Right of Preemption could then be seen more as a safety net against price slippage than as a device to be used automatically. The publicity concerning creation of the zone could be a key factor in this respect. 4. Land taxes and their use as mobilization tools 4.1 The taxes and duties assessed on land and real property (urban tax, taxe d'ddilit6, business license fees [patente], urban land tax, real-estate profits tax) are highly diverse, complex, unstable, unproductive, and costly to manage by comparison with other taxes. The system is particularly unstable in that it reflects the various priorities assigned over the years to different objectives: efforts to control speculation in 1978, through the creation of two new taxes, which were abolished in 1983, encouragement of real estate investment in 1981 through the exemptions contained in the Real Estate Investment Code, redistribution of tax proceeds to benefit the local authorities (1989), etc. This instability has reduced the transparency of Morocco's land and real-estate tax system. 4.2 The transaction tax (recording fee) was very high (average of 17%) before the 1978 tax reform and the Real Estate Investment Code of 1980. The present rate under ordinary law is set at 5%. The Real Estate Investment Code introduced a lower rate of 2.5% for purchases by professionals of building sites or of buildings slated for reconstruction, subject to a commitment by the purchaser to build within four years. This reduction in rates had a marked effect in reinvigorating real estate activity in the ensuing years. Given that market prices are systematically underreported at the time of sale, the present level of recording fees, while it can still be reduced, should be seen as an adequate approximation of the administrative costs of real- estate transactions. 4.3 The urban tax, an annual tax assessed on apartment building value, represents a substantial share of local tax revenue (DH 279 million in 1993). Its amount is steadily decreasing, however, and it presents a number of difficulties in terms of computation (the five- year revision of rental values does not follow market price trends), high management cost, and the scale of the disputes to which it can give rise. 4.4 The need for action on the land markets was reflected in the 1978 Finance Law for 1978, which created three new taxes: the tax on real-estate profits, the single tax on the Annex D Page 10 of 12 inclusion of sites in urban perimeters, and the tax on unconstructed sites. Given the inadequacy and the poor performance of these taxes, it was decided to merge them in the form of a o1ta gains tax. 4.5 The tax on real-estate profits, created in 1978, is applied to the capital gain arising when an apartment building is sold or a title deed transferred. Following a number of changes in the rates and terms of this tax (computation of purchase price, various deductions, exemptions), the rate is presently fixed at 15% of the capital gain with a threshold set at 2% of the transfer price. Because of systematic underreporting at the time of sale, the complexity of the mode of computation of this tax, and the absence of a time frame for its payment, its yield has so far been mediocre. 4.6 The tax on raw land, created by the 1978 Finance Law, applied in 1981 and 1982, repealed in 1983, then reintroduced in 1989, is geared strictly to promoting land supply. It is applied to sites meeting the following conditions: - included in an urban perimeter - provided with water and electricity hookups - not affected by a building prohibition - purchased over three years ago. As a means of encouraging the placing of land tracts on the market, the rate of this tax increases with the time the land is held: from 1.5% of the market price to 5.5% after five years. In addition to the above-mentioned difficulties of computation and collection, this tax discouraging land withholding complicates the determination of the tax base, leads to a large number of disputes, and produces an insignificant yield. In the 1981-82 version, the law would determine a rate of the tax which increased as long as the land was held: from 1.5% of the market price to 5.5% after five years. The tax was unsuitable in meeting its objectives due to its complexity. The actual law imposes a fixed rate of tax which depends on the time of land use: (DH/m2 for villas zoning and 4 DH/m2 for apartment building zoning). This value is very low compared with the market price and the tax does not create any incentive while the yield remains low. Foreign experiments with this type of incentive tax have generally ended in failure. 4.7 The third component of the 1978 reform, similar in its intention to the tax on unconstructed tracts, was the single tax, charged at the time of sale of farm lands recently included in the urban perimeter, and also charged in the event of a zoning change (from a "villa zone," for example, to an "apartment building zone"). This tax, imposed at the flat rate of 30% on the capital gain arising from the zoning change, was never applied, owing to technical difficulties, and was revoked by the 1983 Finance Law. Annex D Page 11 of 12 4.8 Given this satisfactory situation, there is now a proposed draft law on the land capital gains tax, designed to replace the 1978 legislation. It includes two changes: - the new tax will be collected by the local authorities within the framework of the local taxation system; - the proceeds of the tax will be allocated to a "National Land Fund," whose procedures have yet to be outlined, but whose strict purpose would be to finance the purchase of public land. Several comments are in order here, reflecting a certain skepticism concerning implementation of this proposal: the technical complexity already mentioned, in particular the choice of the tax-generating event (first sale following the zoning change), will produce the same effects (difficulties of computation, disputes, difficulties of collection). The collection mechanism demands perfect coordination between the communes (or urban agencies), the notaries and the adouls, who must be notified of the tax collection perimeter, and the tax authorities, in particular the recorder's office. Such coordination does not operate today. Lastly, this type of tax is a clear incentive for owners to hold onto their land, to underreport the sale price, and to avoid recording transactions. Given all of the above factors, the proposal in its present form should be abandoned and replaced by an annual land tax on market value. 4.9 Since the tax collected at the time of sale always gives rise to the same problems (withholding of land, under-computation, failure to record, disputes), it is suggested that an annual general land tax be put in place for all non-agricultural lands, regardless of the zoning plan, based on market value. The introduction of such a tax, which would not only be productive but would also serve as an incentive, should succeed in making the land markets more transparent, first through the development of land value observation centers based on the opinion of experts (i.e. the public authorities, real estate agents, public and private promoters), the production of land price charts, and finally the publicizing of deeds of transfer pertaining to all land and real estate transactions. Simulations would need to be carried out, based on observation center data, to determine the orders of magnitude for the possible yield of such a tax. Annex D Page 12 of 12 Références Ministère de l'Intérieur Colloque national de l'habitat, Rabat, 27-28 octobre 1994 Recommandations de la Commission No. 1 "promotion immobilière et marché foncier de l'habitat" Programme de Construction de 200 000 logements Rapport de la Commission Foncière US-AID-PADCO Etude de la politique foncière urbaine au Maroc Juillet 1987 Crédit Immobilier et Hôtelier Etude du Secteur de la Construction 6 volumes, INAU-CERAU et TUM, Rabat, décembre 1992 Banque Mondiale Gestion Foncière au Maroc et en Tunisie Actes du Symposium de Barcelone, 22-25 septembre 1986 Abdellah Lehzam Le logement urbain au Maroc CCMLA, Rabat, mars 1994 Association Nationale des Architectes et Urbanistes US-AID-Bureau régional pour l'habitat et l'urbanisme Stratégie urbaine et urbanisme opérationnel Actes du séminaire tenu à Rabat, du 11 au 13 mai 1989 Ministère de l'Habitat Direction de la Planification et de la Programmation Etude relative au processus de production et de développement du secteur de l'habitat au Maroc 7 volumes, Team Maroc / Team International, Rabat, octobre 1994 MG/sg m:\sylvie\nouveau c:\j\mor\report\houstrat\annex-c Annexe E Page 1 of 17 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Visible and Less Visible Subsidies David Sewell EMTIN September 26, 1994 Annexe E Page 2 of 17 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Visible and Less Visible Subsidies TABLE OF CONTENTS Introduction ............................................... 4 Subsidies for Mortgage InterestRates ............................... 4 Income Tax Exemptions ....................................... 5 Exemptions from Capital Gains Taxation (TPI) ......................... 6 Exemptions from Registration Taxes and the Taxe Urbaine ................. 6 Overtaxation of Real Estate: TVA on Interest ..................... 7 Sum m ary ............................................ 7 Identifying Implicit Special Tax Treatment: Methodology ................... 7 Effective Tax Rates. ..................................... 7 Tax Expenditures. ...................................... 8 Income Tax Treatment ............................... 8 Consumption Tax Treatment ............................ 9 Summary: Benchmarks for Assessing Tax Expenditures in Owner-Occupied Housing and otherHousing ............................ 9 Non-owner OccupiedHousing. .......................... 9 The Value of Special Tax Treatment for Non-owner Occupied Housing .......... 10 The Value of Mortgage Interest Deductibility .......................... 11I Distributional Effects of Mortgage Deductibility ......................... 11 Conclusions ............................................... 13 Annexe E Page 3 of 17 KINGDOM OF MOROCCO HOUSING SECTOR STRATEGY Visible and Less Visible Subsidies Introduction Our purpose is to estimate the value of subsidies and less visible incentives or examples of special treatment for housing in Morocco. It is relatively simple to determine the value of overt subsidies such as the interest rate subsidies that are given in Morocco, and one can identify and proceed to calculate the value of explicit tax exemptions. In not all cases is such special tax treatment identified as such, however, and we shall spend some time explaining the methodology used to derive our estimates of implicit "tax expenditures" In addition to our own estimates, we shall use estimates provided by the Moroccan Ministry of Finance and the IMF for the value of some of these housing incentives. Not all of these estimates are for the same year. Since it was not possible to redo all of these estimates on a common base, we have simply adjusted some by means of price indices. Our conclusions will combine some comments on the estimates with some additional remarks on the microeconomic implications of our findings. Subsidies for Mortgage Interest Rates The most visible subsidies for housing in Morocco consist of the percentage interest rate reductions in mortgage terms, which had a total value of DH190 million in 1992. Low-income housing receives the most generous terms: a 5 percentage point reduction in mortgage interest rates was given in 1992 where the mortgagee's income did not exceed DH36,000 per annum, the size of the housing unit did not exceed 100 square meters and the value of the housing did not exceed DH150,000. More than two-thirds of total mortgage interest subsidies in 1992--68%--were distributed for such low-income housing. In 1992, a 2 percentage point reduction in mortgage interest rates was given where the value of the housing unit did not exceed DH300,000 and the purchaser intended to live in or rent the unit. This general housing program accounted for a further 26% of the value of all mortgage subsidies in 1992. Finally, the remaining 6% of total mortgage subsidies in 1992 were given to housing being purchased in Morocco by Moroccans residing abroad. The latter benefited from either a 4 percentage point reduction or a 2 percentage point reduction in mortgage interest rates depending on whether the value of the mortgage was under or over DH100,000 respectively, no other conditions being applied with respect to either the incomes of the mortgagees or the value or size of the housing unit being purchased. Annexe E Page 4 of 17 Income Tax Exemptions The Building Investment Code (Code Immobilier) provides a 5 year exemption for real estate promoters from the corporate income tax after construction is completed (for 50% or 100% of income taxes according to the location of the city, with activity in the principal cities receiving the less favored treatment), and a 3 year 100% exemption from income taxes for individuals constructing buildings for rent. According to calculations made by the IMF and the Ministry of Finance, exemptions of both personal and corporate income taxes in 1990 in this Code accounted for a total estimated value of 120 million DH of lost tax revenues. This income tax exemption of 120 million DH was equal to 22% of the value of such income tax exemptions in all the Investment Codes apart from that for agriculture. Income in agriculture is totally exempt from taxation until the year 2020. Considering only the non- agricultural investment codes, the income tax exemptions in the Building Investment Code were second in importance only to those for export industries, which accounted for 51 % of total exemptions, but were more valuable than those for other domestic industries, which accounted for 18% of total income tax exemptions given by the investment codes. Tax exemptions from housing would be less than the 120 million DH calculated for all real estate in 1990. Housing accounted for 84% of the total value of construction in that year.' On the other hand, the consumer price index was 14.2% higher in the base year for our estimates, 1992, than in 1990. Taking both these factors into account, the estimated value of the income tax exemptions in the Code Immobilier would have been roughly 116 million DH in 1992. Restrictions were imposed in 1992 on use of the income tax exemptions in the Building Investment Code. For instance, in the case of the 5 year exemption from all or half of corporate and personal income taxes, a minimum payment equal to 25 % of the amount that would otherwise be payable is required.' Further, although the tax on capital gains in real estate is 15%, a minimum tax equal to 2% of the sale price is applied. These minimum tax provisions in the corporate and personal income taxes came into effect in 1992, and would therefore apply for the full taxation year of 1993. One result of these attempts to limit the tax exemptions in real estate by indirect or roundabout methods is a system of taxation whose complexities raises costs of taxpayer compliance and administration. A more straightforward and transparent system of real estate taxation would have its advantages. 1/ Annuaire Statistique du Maroc 1991, p.180. / One implication is that it should be relatively easy to calculate the true value of the exemptions in future years; it will simply be a matter of multiplying the value of the minimum taxes paid by four. Annexe E Page 5 of 17 Exemptions from Capital Gains Taxation (TPI) Under the Building Investment Code, capital gains from the first sale of a building are exempted from capital gains taxation for 5 years after a permit for construction is obtained. The exemption is for 100% of capital gains if the building is located outside the larger cities and for 50% if the building is located in the larger cities. The Ministry of Finance has estimated the value of such exempt capital gains for 1991 for the purposes of our enquiry as being 747.8 million DH. Several adjustments need to be made to this figure for the purposes of our enquiry. First, it is for all construction, only 84% of which was made up of housing in 1991. Second, since the consumer price index increased by 5.7% between 1991 and 1992, the value of capital gains needs to be accordingly adjusted to find the value in the base year of our enquiry. Finally, if the capital gains concerned had been taxable, a tax rate of 15% would have applied and the estimated value of the exemption needs to exclude this amount. The indicated manipulations to our data leave us with an estimated value for the exemption from capital gains taxation of 564.4 million DH in 1992. Exemptions from Registration Taxes and the Taxe Urbaine Special treatment in the income tax is not the only case of special tax treatment for real estate in the Building Investment Code. Half or full exemptions are also given for registration taxes (droits d'enregistrements) and for a period of 5 years following construction from the taxe urbaine. We have not attempted to assign a value to relief from registration taxes since there is some question as to whether exemptions from this tax are not more "standard" than payment of the nominal "full" tariff. Registration duties amounting to 5% of the price of property are theoretically able to be levied, although exemptions of 50% and 100% are available for housing, leasing and land purchases subject to the Building Investment Code and other investment codes. The taxe urbaine is the principal tax on the rental value of real estate in Morocco. Business users have to pay a flat rate of 13.5% of the rental value of their premises, although owner-occupied housing obtains a reduction of 75% of its rental value and then pays according to a progressive rate schedule above a threshold level. The Building Investment Code gives new buildings a 5 year exemption of payment from the taxe urbaine following construction, and low- income housing gets an additional exemption during the period when construction loans are being repaid. It is relatively simple to calculate the value of the exemption from the taxe urbaine, since the taxe d'Wdilit6 is also levied on the rental value of properties and there are no exemptions from the latter tax. Given the relatively low taxes on rental value that have to be paid for housing, however, it is perhaps to be expected that the exemptions from the taxe urbaine in the Building Investment Code are not of high value. For the purposes of our enquiry, the Ministry of Finance has estimated the value of these exemptions for housing in 1991 as being worth 22.3 Annexe E Page 6 of 17 million DH. Adjusting for the change in the consumer price index of 5.7% between 1991 and the base year for our estimates, 1992, the value in 1992 DH would have been 23.6 million DH. Overtaxation of Real Estate: TVA on Interest. It needs to be recognized that in at least one respect, real estate income is or may well be overtaxed by standards elsewhere. This is with respect to the VAT that is applied to interest payments. the VAT is intended to apply to added value in transactions, but in the case of financial transactions in Morocco is applied to gross interest payments. This is important, given Morocco's high real and nominal interest rates on mortgages and other loans. In fact, most countries do not apply the VAT at all to financial transactions. In 1992, the government reduced the VAT to 7% for housing and required it to be paid only for housing valued at more than 500,000 DH. While this tax is now less important, it would still be relevant to calculate a "negative "tax expenditure for such interest payments. Summary Table 1 summarizes our estimates of the various explicit exemptions from taxation of housing, which sum to 894.9 million DH for 1992. As can be seen from Table 1, these estimates are dominated by the exemptions from capital gains taxation. Identifying Implicit Special Tax Treatment: Methodology Effective Tax Rates. There are several methods of identifying cases of implicit tax incentives or special tax treatment. A method we have used elsewhere is to compare effective tax rates on investment in particular assets.' If the tax system is neutral between alternative investments, they should bear the same tax rate. Our previous investigations did demonstrate that investments in land and buildings in Morocco are, in fact, taxed less heavily than investments in the other principal asset classifications-- equipment and inventories. This is because land and buildings have longer effective lives than the other assets and interest costs involved in their acquisition are deductible over longer periods than for shorter-lived assets. In fact, with 100% debt financing and assuming 5% inflation, we have shown that effective tax rates are actually negative for investments in land.4 3/ See Jack Mintz, David Sewell and Tom Tsiopoulos, Tax Effects on Investment in Morocco EMPTM, 1994. 4/ Ii., pp.18-21. Annexe E Page 7 of 17 Unfortunately, time did not permit measurement of effective tax rates for investment in housing. Tax Expenditures. An alternative and time-honored approach to valuing implicit tax incentives and other forms of special treatment is to estimate "tax-expenditures". This phrase is meant to indicate that expenditure programs can in effect be delivered through the tax system, by means of departures from normal or "benchmark" taxes. Normative judgements enter into the choice of the benchmark system with which actual tax treatment is to be compared, however, so such estimates of tax expenditures are not value-free. There are, in fact, two well-known although competing normative benchmarks which can be used to assess tax expenditures in Moroccan housing: the benchmarks of income and consumption taxes. Income Tax Treatment. If housing were to be treated like any other investment in an income tax system, borrowing costs would be deductible and all income from the asset would be taxable. The Moroccan treatment of housing departs from this standard. Mortgage interest costs are deductible from taxable income, but capital gains from sale of owner-occupied housing are not taxable if the latter is held for at least 8 years. Other capital gains realized from real estate (such as the sale of housing which is not a principal residence) is taxed at a preferential rate of 15% after inflation. It should be recognized that home ownership is fostered in most industrial countries as an objective of social policy and most such countries also treat owner-occupied housing more generously than the treatment this asset would receive in an income tax that was neutral in its treatment of alternative investments.' In particular, a true income tax would involve taxing the imputed rental income of owner-occupied housing. Morocco is not alone in ignoring the value of such returns from housing; in fact, few countries attempt to do so. Indeed, the U.S., Canada and some other countries deliberately choose not to record non-taxation of such implicit income in their tax expenditure tables. 5/ An additional reason for the use of effective tax rates is that the use of incentives is legion in Morocco's tax system. Where special treatment is extended to many, in effect nobody may get a tax break. Nominal tax rates lose their meaning and it takes a comprehensive analysis such as the calculation of effective tax rates to reveal actual comparative tax burdens. 6/ Home-ownership is, in fact, the primary asset of the average family even in the richest economies. 7/ The first tax expenditure table issued by the government of Canada in 1979 treated non-taxation of the imputed rental value of owner-occupied housing as a tax expenditure and indeed it was found to be the largest tax expenditure in the personal tax system. Soon after the publication of this table, the then Minister of Finance had to issue a denial that he intended to tax such imputed rental values. In subsequent tables, the item was dropped as a tax expenditure on the grounds that there is a political consensus against such taxation. For the same reason, the U.S. tax expenditure accounts have always excluded this item. Annexe E Page 8 of 17 Consumption Tax Treatment. If housing were to be assessed against the benchmark for a consumption tax, a different set of criteria would be employed to determine tax expenditures. Under a consumption tax, those investing in owner-occupied housing would not be allowed to deduct the costs of such investments (such as mortgage interest) from taxable income as they would in a true income tax. On the other hand, there would be no tax on the yield from housing under a consumption tax, either in the form of the implicit rental value of the services from such housing or any capital gains when the asset is sold.! Summary: Benchmarks for Assessing Tax Expenditures in Owner-Occupied Housing and other Housing. We choose to evaluate tax expenditures for owner-occupied housing in Morocco against a consumption tax standard. Our reasons are that taxes on personal income pr s amounted to only 15% of Moroccan tax revenues in 1992 and were dwarfed in importance by taxes on consumption: the VAT alone accounted for 25% of total tax revenues. Further, some taxes on income are really flat taxes which do not vary with the income of the taxpayer (e.g., taxes on interest income) as specified in the rate structure of the personal income tax (IGR). These factors lead us to categorize the Moroccan tax system as being more consumption-based than income-based. The principal tax expenditure for owner-occupied housing under a consumption tax base would therefore be the deductibility of mortgage interest, since under a consumption tax the revenues from owner-occupied housing would not be, and indeed are not, taxed in Morocco. It is of interest that deductibility of mortgage interest is nowhere mentioned as a form of implicit subsidy in the available literature on housing in Morocco. Non-owner Occupied Housing. Tax expenditures in investments in housing other than that occupied by owners will be assessed with reference to the tax treatment applied to other investments or sources of income. The principal example of special treatment for real estate in Morocco consists in the fact that returns are taxed at a rate of 15% after inflation. Such taxation of "real" returns does not in itself constitute a tax expenditure: indeed most proponents of an income tax would think it is only appropriate to tax real and not inflationary income. The fact that the resulting tax rates for investments in real estate are so much lower than those for other assets does constitute special treatment, however. 8/ The consumption tax treatment of housing that we have described is the so-called "prepaid" version of a consumption tax. It is the same principle as is applied in a VAT, which permits business to subtract the cost of inputs and capital goods from the value of output. See David Bradford, Untangling the Income Tax, (Cambridge, Mass., Harvard University Press, 1986), p. 85. Annexe E Page 9 of 17 The Value of Special Tax Treatment for Non-owner Occupied Housing. A first approximation of the difference in tax treatment between non-owner occupied housing and other investments can be obtained by adding 1992's inflation rate of just under 5% to the tax rates on various sources of income (other than capital gains taxation) in that year. The latter were a top personal marginal tax rate of 52%; a marginal tax rate on dividend income of 52.4%; and tax rates on interest income of 20 and 30% respectively.' The 15% tax on real estate income compares extremely favorably with the resulting tax rates, taking into account inflation, of 56.9% for wages, 57.3% for dividend income and 24.9% or 34.9% for interest income in 1992. Total capital gains taxation on buildings paid in 1992 was 682 million DH, and as noted the rate of the tax is 15%. This implies that total profit in real estate declared for tax purposes in 1992 was 4549 million DH. It is not possible to estimate capital gains tax originating separately in housing and other building activities. We have therefore assumed that capital gains tax arising from housing is proportional to the value of housing construction in total housing, industrial and commercial construction in 1992, or 88%0 On this basis, as Table 2 shows, tax actually paid on capital gains in housing is estimated to have been just over 600 million DH in 1992. The taxpayer's lowest conceptual opportunity cost, in terms of taxes of taxes paid on other sources of wage or investment income, would have been to have this income taxed at the rates of taxation for interest, which would have yielded taxes of 997 million DH or 1397 million DH. Alternatively, taxation at the highest tax rate bracket of the personal income tax, which would seem likely to be the relevant tax bracket for those receiving real estate capital gains, would have cost the taxpayer 2278 million DH. Finally, taxation of real estate profits at the same rate as dividend income would have yielded 2294 million DH in 1992. Another way of putting these estimates of capital gains tax paid on housing and the tax expenditure element in capital gains tax in context is to compare them with revenues raised by the corporate and personal income taxes (the IS and IGR). Our estimate of actual capital gains tax paid in housing is equal to 4.3% of total revenues from the personal and corporate income taxes in 1992, and we can provide low and high estimates of revenues that would have been raised had this income been taxed at rates applicable to other sources of income. Had real estate capital gains been taxed at the lowest applicable rate for interest income or at the tax rate applicable to corporate dividends received as income, revenues would have 2/ Interest paid to unidentified lenders (e.g., holders of corporate bonds) is subject to a 30% tax which is withheld by the borrower. Where the lender is readily identifiable (as for bank deposits), interest is taxed at a 20% rate which is withheld by the borrower. This withholding tax is then credited against total personal income tax liabilities. 10/ See Annuaire Statistique du Maroc 1993, p. 191. Annexe E Page 10 of 17 been respectively equal to 7.2% and 16.6% of total personal and corporate income tax paid in 1992. The Value of Mortgage Interest Deductibility Some institutional features of money markets in Morocco help explain why deductions of mortgage interest could be an important tax expenditure. Interest rates in capital markets in Morocco are high, largely because of preferential access of government to loanable funds. In 1992, short-term interest rates were around 14%. Credit Immobilier et Hotelier (CIH) is the main mortgage lender in Morocco and about the only mortgage lender in urban areas, although there are programs for rural borrowers. Private banks sometimes provide medium-term mortgage financing to their affluent customers as an accommodation but do not generally make long term mortgage loans." In 1992, CIH was supplying mortgage funds at rates of 13% and even with subsidies as great as 5% in some programs aimed at low income groups, relief from interest costs was clearly important an important feature of the tax system. Table 3 shows interest rate subsidies for various categories of housing in 1992. Low- income housing (HBM) received an interest rate reduction of 5%, general housing (r6gime g6n6rale) received an interest rate reduction of 2% up to a value of housing of 300,000 DH and housing in Morocco for Moroccans working abroad received an interest rate reduction of 2% for approximately 70% of the program outlays and 4% for the remaining 30% of program outlays.12 Corresponding amounts of mortgage interest which are deductible for income tax purposes, calculated as the interest rate of 13% on mortgages minus the appropriate interest rate subsidy, are shown in Table 3. Distributional Effects of Mortgage Deductibility Although the primary purpose of this report is to outline and quantify visible and implicit subsidy elements in Morocco's tax and expenditure systems for housing, the distributional effects of the various measures are of some interest and are indeed required to calculate the value of mortgage interest deductibility. Many of these programs have specific redistributional objectives of aiding low-income groups. In the case of tax expenditures such as the deductibility of mortgage interest from taxable income, however, the distributional effects may be less evident and may, indeed, be very surprising to some. We shall attempt to indicate the distributional effects of mortgage interest deductibility. 11/ See MNIIN Staff Appraisal Report, Land Development Project for Low Income Families, 11693- MOR, 1993, pp.5 and 8. 12/ The rural subsidy program (habitat rural) was not in effect in 1992. In making our calculations of the tax expenditure for mortgage interest paid by Moroccans working abroad, we assumed alternatively that the income distribution of this population is the same as that of the populations receiving mortgages under either the general program or the low income program, depending on the size of the interest rate subsidy received by the foreign workers. Annexe E Page 11 of 17 Table 4 provides data on the income distribution of CIH clients obtaining mortgages. The table also shows the tax rate for each additional dirham of income for income classes corresponding to the distribution of those receiving mortgages. This marginal tax rate can alternatively be viewed as the amount of assistance those with mortgages receive for each additional dirham of interest paid." A distributional feature of mortgage interest deductibility which is immediately evident from Table 4 is that its benefits are confined to those who pay taxes. Although CIH's special regime is designed to provide mortgages to low-income people, 29 % of the participants in the program receive no benefits at all from mortgage interest deductibility because their incomes are so low that they are exempt from paying taxes. A second effect of the deductibility of mortgage interest evident in Table 4 is that its benefits increase with income level. Those in the top personal income tax bracket in 1992 in effect had 52% of the interest costs for their mortgages paid for by the government. This is an "open-ended" subsidy, too, since there is no limit on the amount of mortgage interest that is deductible from income tax owed. In terms of overall distributional effects, about 8% of mortgage holders in CIH's general program and all of the mortgage holders in their special program for low-income people received assistance equal to 22% of their mortgage costs. Some 62% of those receiving mortgages in CIH's general program receive at least twice this level of assistance since they are able to deduct mortgage interest costs at the marginal tax rate of 44% or more. The kinds of conclusions that we have made above are simply based on comparisons of the income distribution of the relevant populations receiving mortgages and their marginal tax rates. Examination of the actual estimates of tax expenditures in Table 4 reinforces these conclusions. Clearly, the largest tax expenditures from mortgage deductibility occur in the general program. It can be seen in Table 4 that tax expenditures for the general program are 4.6 times those for the low-income program, although the amount deductible from income tax is shown in Table 3 to be only 1.3 times that in the low-income program." Put another way, tax expenditures for mortgage deductibility are 41% of the amount deductible from income tax in the general program but only 11.9% of the amount deductible from income tax in the low- income program. 13/ There are some biases in this procedure which should be outlined. First the tax rates in Table 4 are actually those for taxable income, i.e., after deductions are permitted for dependents, etc. The income distribution for those having mortgages in Table 4, on the other hand, is for total income of the family. Comparisons between the two series therefore biases upwards data in Table 4 as a measure of the marginal tax rates faced by mortgage holders. An offsetting bias is that we are ignoring mortgage finance or medium-term finance for housing purposes for those with high incomes that is provided by the banks and other financial institutions than CIH and whose interest is also deductible from income tax. The net effects of these two factors is unknown. 14/ These and all following comparisons in this section ignore the program for foreign workers. Annexe E Page 12 of 17 Finally, it is worth noting that if the tax expenditures for all mortgage deductibility in the various mortgage programs in 1992 are summed, the total of 155.6 million DH is 81 % of the value of overt interest subsidies of 190.9 million DH shown in Table 3 for the same year. Clearly, tax expenditures cannot be ignored as an element of hidden subsidies. Morocco is, of course, far from being the only country which permits mortgage interest to be deducted from taxable income. The regressive features of mortgage. interest deductibility that we have noted have been called "upside-down equity" by the advocates of tax expenditure tables." They add that it is extremely unlikely that an expenditure program would be proposed which offered similar substantial rates of subsidy to high- income groups and no subsidy at all to those too poor to pay any income taxes. Conclusions Our calculations of the value of explicit and implicit subsidies for housing reveal that the former are dwarfed by the latter, and visible subsidies in turn are dwarfed by explicit tax exemptions. The most visible subsidy element is that of interest rate subsidies, which amounted to 191 million DH in 1992. We calculated that the total value of explicit subsidies and tax exemptions was 895 million DHMin 1992, with capital gains exemptions accounting for 63% of total tax exemptions.. Turning to tax expenditures, we found that the value of mortgage interest deductibility was equal to 81 % of mortgage interest rate subsidies and benefits to participants in the general mortgage program were 4.6 times those for participants in the low-income mortgage programs. We suggested that it is unlikely that a direct expenditure program with these regressive characteristics could be suggested, defended and implemented. All other elements of tax expenditures and explicit subsidies are puny, however, when compared to some of the estimates of taxes that would have to be paid if housing bore the same tax rates as other investments. We calculated that if capital gains on housing were taxed at rates applicable to interest and dividend income, for instance, an additional 400-1700 million DH would have been owing in 1992. Comparing housing with other sectors, there is also little doubt that it receives explicit and implicit subsidies and tax expenditures that are only exceeded by those accruing in the agricultural sector. Despite these incentives, Morocco still has a housing problem and we might venture to guess that the net result of the various incentives for building may have been to confer additional rents on a factor already in scarce supply-- the limited amount of urban land. 15/ The classic statement is in Stanley S. Surrey, Pathways to Tax Reform, (Cambridge, Mass., Harvard University Press, 1973), pp. 36-37. Annexe E Page 13 of 17 Finally, taxation of the real estate sector shares the same characteristic as much of the Moroccan tax system--it is overly complex. The numerous tax incentives and examples of special treatment raise both administrative and compliance costs. Even attempts to limit the value of these exemptions, such as the minimum taxes that were introduced in 1992, have added to this complexity., This is not to say that progress has been absent in the tax and expenditure treatment of this sector. Subsidies have been limited, and to enhance competition between public and private developers, the former are no longer exempted from income taxes. Further, there have been some cases of inequitable and inefficient of tax treatment in the sector, such as the VAT on interest payments, whose effects have been ameliorated. This tax has been reduced in level from 14% to 7% and its scope has been narrowed to relatively high income housing. We have no doubt that it will soon be removed altogether. Annexe E Page 14 of 17 Table F-1: The Value of Visible Subsidies and Tax Exemptions for Housing, 1992 million DH Interest Subsidies 190.9 Income Tax Exemptions 116.0 Capital Gains Tax Exemptions 564.4 Exemption from the Taxe Urbaine 23.6 TOTAL 894.9 Annexe E Page 15 of 17 Table F-2: Tax Expenditures in Capital Gains Tax on Building (the "Taxe sur les Profits Imnobiliers"), 1992 Tax Rates Taxable Capital Gains Tax Expenditures 1) 2 3) 4) 5) 6) 7) Nominal )Where Applied Annual Total Real Taxable capital Taxable Capital Opportunity Cost of Rate (%) Rate of Tax Rate Gains: Housing, Gains in Present Taxation of Inflation (Cols 1+3) Industrial and Housing (88% Capital Gains in 1992 (%) Commercial Real of Col 5) Housing (Col 6 (%) Estate (Col. 4* minus present taxable 4549 million gains in housing, or DH,or estimated 600 million DH) actual real estate profit in 1992) 15 TPI 15 682 600 (real rate) 52 IGR, top rate 4.9 56.9 2588 2278 1678 bracket 52.4 Tax on dividend 4.9 57.3 2607 2294 1694 income 20 Tax rate on 4.9 24.9 1133 997 397 interest income whose source is declared 30 Tax rate on 4.9 34.9 1587 1397 797 interest income whose source is not declared Notes: 1) Actual taxable profit in real estate is estimated to have been 4549 million DH in 1992. 2) Nominal tax rates on dividend income in 1993 are derived by the method used in Jack Mintz, David Sewell, and Tom Tsiopoulos, Tax Effects on Investment in Morocco pp.29, but substituting the corporate tax rate of 40% (44% including the PSN) which applied in 1992 in the calculations. Annexe E Page 16 of 17 Table F-3: Value of Interest Rate Subsidies and Income Tax Deductions for Housing Low Income Housing General Housing Housing for Moroccans Total (HBM) (Regime Generale) Working Abroad (TME) Subsidy Rate 5% 2% 4% for 30% of loans 2% for 70% of loans Subsidy Value 129,597,592 50,402,305 5,011,453 (4%) 190,894,359 5,883,009 (2%) Rate Deductible from 8% 11% 9% for 30% of loans - Income Tax 11% for 70% of loans Amount Deductible from 207,356,147 277,212,678 11,275,769 (9%) 528,201,144 Income Tax 32,356,550 (11%) Annexe E Page 17 of 17 Table 4: Distribution of Tax Expenditures for Mortgage Recipients by Income Class and Marginal Tax Rates, Morocco, 1992 General Program (RG) Low-Income Program (HBM) and Low-Income Foreign Worker (TME) Program Annual Distribution Marginal Tax Foreign Worker: Distribution of Tax Low-Income Income of Income: Rate Expenditure: General Program Population by Expenditures: Foreign General General (TME) Income Class Low-income Worker Program Program Program Program (RG) (HBM) (TME) 0- 12,000 0.4 0 - - 29 - 12,001 - 24,000 0.4' 14 155,239 18120 46' 13,353,736 726,160 24,001 - 36,000 7.2 22 4,391,049 512,528 25' 11,404,588 620,167 36,001 - 60,000 30 36 29,938,696 3,494,507 60,001 - 90,000 20.5' 44 25,004,584 2,918,561 90,001 - 120,000 20.5' 46 26,141,156 3,051,223 120,001 - 200,000 10.5' 48 13,971,519 1,630,770 > 200,000 10.5' 52 15,135,812 1,766,668 Total 114,738,328 13,392,377 24,758,324 1,346,327 Source: Text and Credit Immobilier et Hotelier, Rapport du Conseil d'administration sur I'Exercice 1992, p. 44 I/ Totals interpolated between adjacent income brackets 2/ Tax Expenditures are calculated as a program's share of population in an income class times the marginal tax rate for that income class times the amount deductible for income tax for the program as shown in the last line of Table 3. c:\j\mor\report\houstrat\annex-f.eng Annex F Page 1 of 3 BIBLIOGRAPHY WORLD BANK PUBLICATIONS Kingdom of Morocco, Poverty, Adjustment and Growth, January 1994. Kingdom of Morocco,Public expenditure : issues and outlook, August 1994. Kingdom of Morocco, Fifth Water Sullpy Project, SAR October 1993. Financement du Logement en Tunisie, 1994, Credit Agricole consultants. Housing, Enabling markets to Work, 1993, a World Bank Policy Paper. The Housing Indicators Program, UNCHS and the World Bank, april 1993. Bertaud Alain, Renaud Bertrand, 1993, Cities without land markets, discussion papers. Farvacque Catherine, Mc Auslan Patrick, 1993, Reforming Urban Land Policies and Institutions in Developing Countries, Urban Management Programme, Paper no 5. Vittas Dimitri, Iglesias Augusto, 1992, The rationale and performance of personal Pension plan in Chile, working paper. IFM/MIGA Foreign Investment Promotion in Morocco, Plan Location International/Price Water House. Lea, Michael J. and Bertrand Renaud, Contract Savings for Housing - Suitability to TSE Financial Reforms Tunisia - Strategic Issues in Housing Finance and the Land Market, November 1994. Morocco - Informal Housing - Upgrading and Prevention Policies and Programs, December 1983. MOROCCAN PUBLICATIONS Cr6dit Immobilier et Hotelier, 1992, Etude du Secteur de la Construction, INAU-TUM. Cr6dit Immobilier et Hotelier, Rapport d'activiti de l'exercice 1992. Dryef M'Hammed, 1993, Urbanisation et Droit de l'Urbanisme au Maroc, CNRS Editions. Annex F Page 2 of 3 El Moukhtari Ahmed, cabinet d'audit et d'organisation, 1993, Les organismes publics d'Habitat, valeurs d'exploitation et équilibre financier 1991. Lezham Abdellah, 1994, Le Logement Urbain au Maroc. Kingdom of Morocco, Ministry of Housing, 1991 and 1992, L'Habitat en Chiffres. Kingdom of Morocco, Ministry of Housing, July 1994, Premier Programme de Construction de 200.000 Logements. Kingdom of Morocco, Ministry of Finances, 1992, Compte-rendu de l'Activité des entreprises d'assurances et de réassurances et des principaux régimes de prévoyance sociale. Kingdom of Morocco, Caisse de Dépôt et de Gestion, exercice 1992, Rapport d'Activité. Kingdom of Morocco, Ministry of Housing, 1994, Processus de production et de développement du secteur de l'habitat au Maroc, Team Maroc, Team International. Kingdom of Morocco, Ministry of Housing, 1994, Caractéristiques socio-économiques des ménages des quartiers d'habitat clandestin. Kingdom of Morocco, Ministry of Housing, 1994, Caractéristiques socio-économiques des ménages des bidonvilles. MOROCCAN LEGISLATION Code de Procédure Civile, Maroc, 1992, Sochepress-Université. Loi n.01 1-71 du 12 kaada 1391 (30 décembre 1971) instituant un régime de pension civile. Dahir portant loi n. 1-72-184 du 15 joumada Il 1392 (27 juillet 1972) relatif au régime de sécurité sociale (1) te lque modifié par le Dahir portant loi n. 1-77-216 du 20 chaoual 1397 (4 octobre 1977) créant un régime collectif d'allocation de retraite (2) par la loi n. 7-84 (3) et par le décret n. 2-87-738 du 15 ramadan 1408 (2 mai 1988). Dahir portant loi n. 1-77-216 du 20 chaoual 1397 (4 octobre 1977) créant un régime collectif d'allocation de retraite. Dahir n. 1-80-315 du 17 safar 1401 (25 décembre 1980) portant promulgation de la loi n. 6- 79 organisant les rapports contractuels entre les bailleurs et les locataires des locaux d'habitation ou à usage professionnel. Annex F Page 3 of 3 Dahir n. 1-89-205 du 21 joumada 1410 ( 21 d6cembre 1989) portant promulgation de la loi n. 06-89 modifiant et compl6tant la loi n. 011-71 du 12 kaada 1391 ( 30 d6cembre 1971) instituant un r6gime de pension civile. Dahir n. 1-92-7 du 17 hija 1412 (17 juin 1992) portant promulgation de la loi n. 25-90 relative aux lotissements, groupes d;habitation et morcellements. Dahir n. 1-92-31 du 15 hija 1412 ( 17 juin 1992) portant promulgation de la loi n. 12-90 relative A l'Urbanisme. ACADEMIC LITERATURE AND OTHER Anderson John E., 1993, Use-Value Property Assessment: Effects on Land Development, Land Economics. Baken Robert-Jan, Van der Linden Jan, 1992, Land delivery for low-income groups in Thirld World Cities, Avebury. Capozza Denis R., 1994, The risk structure of Land Markets, Journal of Urban Economics. Capozza Denis, and Yuming Lee, September 1994, The Intensity and Timing of Investment: The Case of Land, American Economic Review. Codigo Urbano, Republica de Costa Rica, 1992. Code de l'Urbanisme, France, Litec Edition, 1992. de Soto Hernando, The missing ingredients,150 Economist years. Doebele William A., 1981, Land readjustment, Lexington Books. Engle Robert, Navarro Peter, garson Richard, 1992, On the theory of Growth Control, Journal of Urban Economics 32, 269-283. Gillon Colin, Bonilla Alejandro, 1992,Analysis of a national private scheme: The case of Chile, International Labor Review, Vol 131, No 2. Lahbil Tagemouati Naima, 1991, Dualisme foncier & Fs, Etudes Fonci6res, Paris. Lim Duck-Ho, 1992, The nonneutrality of the Land Value Tax: Impacts on Urban Structure, Journal of Urban Economics 32, 186-194. Pitelis Christos N., 1985, The effects of Life Assurance and Pension Funds on other savings: The Post-War UK Experience, Bulletin of Economic Research 37:3. c:\j\mor\repot\houstrat\anneK-h CATALOGUERS/FILE CONFIDENTIAL Report No: 13930 MOR Type: SR