Cover Image
close this bookNational Experiences with Shelter Delivery for the Poorest Groups (HABITAT, 1994, 140 p.)
close this folderII. HOUSING THE POOR
View the documentA. The case-study countries
View the documentB. National shelter policies
View the documentC. Housing needs
View the documentD. Shelter delivery
View the documentE. Actors and programmes
View the documentF. Financing shelter

A. The case-study countries

The experiences of shelter delivery for the poor in Mexico, Indonesia and India are presented below. The focus is on the following cities: Mexico City, Jalapa, Jakarta, Bandung, Bombay, Calcutta and Delhi. It is evident from table 5 that Mexico is an upper-middle-income country, while Indonesia is a lower-middle-income and India a low-income country. There is, furthermore, a good correlation between the level of economic and social development.

For the purpose of underlining the countries' status as developing countries, the OECD average GNP per capita (at purchasing power parity) is $US 18,000. UNDP (1993), in its categorizing of countries according to their level of “human development”, places Mexico in the high human development group, and Indonesia and India in the middle and low groups respectively. “Absolute urban poverty” is assessed to be 20 per cent of the urban population in Indonesia (rural poverty, 16 per cent). The figures for India are 40 per cent (and 49 per cent). Data are not available for Mexico from the same source. Carmona (1993), estimates that at least 25 per cent of the urban population in Mexico has no or very low incomes, i.e., they are earning one minimum wage or less (one minimum wage was equal to $US 140 monthly income in 1990).

1. Mexico

According to Lombera (1993), research shows that 30 per cent of the Mexican population are extremely poor, i.e., unable to satisfy their nutritional others regarded as too high. By using a wide-ranging definition of basic needs, including cultural and recreational consumption, a very low percentage of the population is not poor. The minimum wage in 1992 constituted only 13 per cent of this “basket of food and services.” The figure a few years back was 38 per cent. Undoubtedly, poverty is rising both in absolute numbers and in the relative sense of an increasing percentage of the population, in rural as well as urban areas. Due probably to rapid urbanization, urban poverty is now increasing more rapidly than rural poverty. Still, the poor constitute a larger proportion of the rural population than they do in urban areas. It may thus be reasonable to assume that the above figure of 30 per cent of food poverty is on the high side for cities. For the present report, it can thus be justified to use 20 per cent of the urban population as an approximate cut-off point for the category “poor”, in a narrow basic needs understanding of the concept.

Table 5. Basic information on the case-study countries

GDP growtha percentage

GNP/capita growtha percentage


GNP/capita PPPb

Gini index

Infant mortality ratec




























a: Average annual growth
b: Purchasing power parity
c: Per 1000 live births

Sources: The World Bank, 1993b; UNDP, 1993.

Table 6. Percentage of population in Mexico with low income

Income level
(minimum wages)





£ 1





1.1 - 2










Note: The source does not have equivalent data on the level of income of the minimum wage through time. An index gives the following information on the declining real value of the minimum wage: 1970 = 100, 1980 = 109, 1988 = 57, 1989 = 50.

Source: Based on Lombera Gonzalez, 1993.

Table 7. Income levels in Mexico, 1990 (economically-active population)

Income level
(minimum wages)

Monthly income

Share of population


Mexico City








No income








< 0.5

< 70







0.5- 1

70 - 140







1.1 -2

141 - 280







2.1 - 3

281 - 420








421 - 700







5 +

701 +







Source: Based on Carmona, 1993.

Real wages in Mexico peaked in 1976. After 1982, the economic depression - combined with restrictive monetarist policies - has eroded the living conditions of working people to pre-1970 levels (Connolly, 1990). The real urban minimum wage declined, on average, by 7.1 per cent annually between 1981 and 1989 (Cardoso and Helwege, 1992). In 1990, 27 per cent of the economically-active population had incomes of one minimum wage or less (see table 7). Those with half a minimum wage or less constituted 14 per cent. Furthermore, the 20 per cent of families lowest on the income distribution scale controlled only 4 per cent of the total income. The shelter problem for the poor in Mexico is growing. In addition to lack of housing, there is a downgrading of housing stock. It is estimated that 45 per cent of the housing have a deficient standard. The prices on building materials are very high. The minimum wage rose 21 times in nominal terms between 1974 and 1985, whereas the price of building materials rose 33 times. The inadequacy of basic services is pronounced. There is insecurity of land tenure and the distribution of subsidies favours the non-poor.

2. Indonesia

The Central Bureau of Statistics in Indonesia calculates that poverty affects 21 per cent of the urban population (Yoewono, 1993). The urban poverty line was in 1990 set at Rp. 20,614 per capita and month. This is a non-wide basic-needs definition. During the 1980s the total number of poor people has been fairly constant, giving a substantial decline in the proportion of the population in poverty, according to official figures (see tables 8 and 9). The low-income groups are estimated to comprise 40 per cent of the population (see table 10). This includes also those without any permanent source of income. For the purpose of this report, it makes sense to define the “poor” as the lowest 20 per cent and the “lower-income group” those between 20 and 40 per cent on the income ladder. The relative size of the categories “poor” and “lower-income” in urban areas in Indonesia and Mexico are thus the same.

Table 8. Official poverty line and number of poor in Indonesia

Urban areas

Rural areas





Monthly income (Rp.)





Population (million)





1980: $US 1 = Rp. 627
1990: $US 1 = Rp. 1,860

Source: Based on Yoewono, 1993.

Table 9. Poverty in Indonesia (percentage)


Urban areas

Rural areas






















Source: Based on Herlianto, 1993.

Table 10. Income levels, Indonesia (1992)

Income group

Percentage of population

Monthly income
(Rp. thousands)



< 100



100 - 449

Lower middle


450- 1,249

Upper middle


1,250 - 3,500



> 3,500

Rp. 2000 = US$ 1

Source: Based on Herlianto, 1993.

The majority of the urban poor in Indonesia live in unserviced so-called kampungs. The Kampung Improvement Programme (KIP) is directed to assist them. The lower-income group, many of whom are the lowest paid government officials, have access to sites-and-services housing. The poor on the other hand, eke out a meagre living through a number of strategies. These are similar for urban slums in most developing countries. They build their shacks on riverbanks, sides of railways, swampy lowlands and pavements. The diets they can afford and the places in which they live, make them prone to most kinds of illness. Yet, the internal variation of living conditions among the poor is great, from utmost destitution to fairly stable livelihoods in community-organized slums.

3. India

In a huge country such as India poverty rates may vary extensively geographically. The proportion of the urban population below a food (balanced diet) poverty line was for instance in Punjab 51 per cent in 1974 (65 per cent in 1964); in Maharashtra 60 per cent (72 per cent in 1964); and in West Bengal 72 per cent (62 per cent in 1964). The overall Indian average was 65 per cent (65 per cent also in 1964) (Rao and Chandrashekar, 1984). According to official figures, there has been a decline in poverty in India generally, from 55 per cent in 1971, to 46 per cent in 1983 and 43 per cent in 1988 (Minhas and others, 1991).

The figure on poverty in urban areas is estimated to be between 30 and 40 per cent of the urban population on average (Ribeiro, 1993). According to Kundu (1993), the Planning Commission is underestimating the level of urban poverty. By aggregating state estimates, he finds that in 1988 the figure was above 35 per cent. There has been a decline from 1971 (46 per cent) to 1988 (38 per cent) of urban poverty proportionally. Pugh (1990) finds that the range in urban poverty in different surveys and studies is from 35 to 45 per cent. The definition of poverty allows only for some essential non-food expenditures. It is thus a narrow basic-needs concept like the one used in this report. Research shows that about 80 per cent of the urban slum population is below the “poverty line.” They use approximately 80 per cent of their income on food alone. Generally, urban households between the 30th and 40th percentiles on the income distribution spend on average 21 per cent on nonfood items. For households in the lower percentiles the figure is lower (Kundu, 1993). The incidence of poverty in urban slums is very high - Indore (68 per cent), Bangalore (73 per cent), Calcutta (89 per cent) and Madras (90 per cent). The figure for Greater Bombay is only 40 per cent. This figure does not include pavement dwellers who are numerous in Bombay. The above data are, unfortunately, from the mid-1980s or earlier. Yet, according to general economic circumstances and trends, substantial changes in the figures are unlikely to have occurred. It should be remembered that the reliability of poverty data is always somewhat questionable. The figures should be taken as indicators of magnitude only.

In sum, the urban poor in India cannot meaningfully be limited to those 20 per cent lowest on the income distribution scale. The discussion on shelter delivery in this report will focus on the 35 per cent with the lowest income, as constituting the poor. This group's range of standard of living will then be comparable to the poor groups (lowest 20 per cent) in Indonesia and Mexico. The lower-income group in India will in this report include those from 35 to 45 per cent on the income scale.

The basic problems of the urban poor in India, as elsewhere, are lack of employment or inadequate income opportunities forcing them into overcrowded and unhygienic areas. These areas lack basic amenities such as sufficient water supply, sanitation facilities and other physical and social infrastructure. The poor are unable to provide themselves with suitable shelter. Housing conditions are today the most visible sign of urban poverty in India, more so than malnutrition and disease.

B. National shelter policies

1. Mexico

The right to adequate shelter was included in Mexico's Federal Constitution in 1983. A new law (Ley Federal de Vivienda), containing aspects of an enabling approach in line with the GSS, was also passed that year. The objectives of the new national shelter strategy are clearly expressed in the Programa Nacional de Vivienda 1990-1994 (Carmona, 1993). The basic principles of the document are:

· Efficiency of the public housing programmes, and extension of the coverage to low-income groups;

· Improved popular participation in shelter provision, and intensified collaboration between public, private and community actors.

The first principle follows the traditional philosophy of the public sector as a provider of housing. The second opens the possibility for the government to assume the role of facilitator.

The shelter-related objectives of the National Development Plan, 1989-1994 include support for the shelter process so that every Mexican family obtains access to adequate housing, while it aims to take advantage of the multiplier effects of shelter delivery, in order to stimulate production and increase employment.

The National Housing Policy has the following specific objectives:

· Modernize the institutional arrangements in the housing sector;
· Concentrate government initiatives to the low-income group;
· Improve the financial mechanisms for public housing programmes;
· Support the process of decentralization;
· Make the distribution of inputs to house-building more efficient.

In 1993 there was a change in the Federal Administration of Housing. The new authority emphasizes the following points:

· Establishment of more flexible and diverse modalities of guarantee for housing loans;

· Promotion of deregulation and simpler construction rules;

· Enactment of legal reforms allowing more flexible and less complex processes of housing production.

The recent formulation of the national shelter strategy is in line with the GSS, and traditional policies are gradually reduced or removed. The GSS seems to have had a positive impact on this change in the housing sector in Mexico.

To underline the reorientation of policy, three specific sub-policies should be mentioned. First, there is a growing role for CBOs and NGOs in the shelter process. The key policy concept in this connection is concertacimeaning social negotiation. Many new CBOs and NGOs have been formed recently in Mexico. There is also an apex organization, a federation of grass-roots groups, CBOs and NGOs representing more than a million people, the Coordinaciacional de Movimiento Urbano Popular (CONAMUP). The main tasks of this organization are to lobby the government on land and housing issues, and to provide financial and technical support to affiliated organizations.

Secondly, an important change in land management in Mexico is the recently passed reform in the National Constitution with respect to ejido lands. This category of land included 99-years leases for peasants. The leases are now transformed into a flexible right facilitating a conversion of such land to urban use and development.

Thirdly, a government programme has been established for the development of 100 medium-sized towns. The aim is to lessen the pressure on larger cities. Shelter programmes in these towns will obtain special benefits to mitigate migration flows to the metropolitan area. The costs of providing urban infrastructure and services will thus be reduced. This will be beneficial for the poor, given that they choose to settle in the towns rather than in the largest cities.

2. Indonesia

Indonesia laid the basis for its fifth Five Year Development Plan (1989-1994), known as Repelita-V, during the International Year of Shelter for the Homeless (1987) and the year that saw the adoption of the GSS by the General Assembly of the United Nations (1988). Housing authorities were then able to develop a new policy and include it in the State National Development Goal. The change in policy is a move from the government as a “provider” to that of an “enabler”. The focus is now on what people can afford and on the role of local authorities together with the private formal and informal sectors and CBOs in shelter provision. Improved building materials, standardization and land regulation and the support of the construction industry have stimulated the participation of the informal sector in housing development. More people are encouraged to build their homes through CBOs and cooperatives and with the assistance of NGOs. The Indonesian Government has endorsed the GSS, yet its Urban Renewal Programme may lead to a removal of a large number of kampungs in Jakarta (UNCHS, 1991d).

The growth of private-sector house-building during 1989-1991 has benefited the better-off people only. Moreover, land prices have increased as a result of this improved activity. A new policy of rental accommodation is included in Repelita-V to benefit the poor. A total of 20,000 units of rental housing are planned. Yet, this is a rather small number, relative to the need. The Government's policy is to leave the responsibility of shelter provision to the people. The role of the Government is mainly to create business and building opportunities, and to stimulate community participation to enable the people to build their dwellings themselves. Programmes have been initiated to enhance the professionalism of housing agencies through education and training. Perum-Perumnas was established to pioneer large-scale housing development in the urban areas. Housing-finance agencies have also been formed.

To improve the shelter conditions of the poor, the Government created the “very simple house”. This type of dwelling is cheaper than the previously designed “simple house”. It has sufficient infrastructure and is built with low-quality materials and is expected to be finished gradually by the beneficiaries. In Repelita-V the Government minimized the subsidy to credit schemes by limiting it to the most needy people only.

3. India

The objective of India's National Housing Policy of 1992 is to create an “enabling environment” for housing activities. People will be assisted in securing affordable shelter. The Government's role is restricted to that of a facilitator in providing access to developed land, building materials, finance and technology (Kundu, 1993). The eighth Five Year Plan (1992-1997) states that housing is essentially a private activity. The need for government intervention to meet the shelter requirements of vulnerable groups is however recognized. In short, the Government in India changed its policy (on paper) from “bull-dozing the slums” to “environmental improvement of slums.” The pavement dwellers were however left out (Bijlani, 1988).

The aims of the new national shelter policy are:

· Setting up an institutional structure for mobilization of resources at a reasonable cost and for the disbursal of funds to the housing sector;

· Creating housing loans as a financial service based on the principle of affordability, recoverability and profitability;

· Restructuring the public agencies for taking up the development of housing sites, particularly for the poor, instead of building dwellings on a large scale;

· Ensuring accountability and efficiency in the public housing agencies by making them increasingly dependent on institutional funds rather than on budgetary support, thereby reducing subsidies;

· Providing subsidized shelter for the poor through both direct public house construction and indirectly by helping them to establish CBOs;

· Removing regulations and administrative barriers impeding the efficient functioning of land and housing markets;

· Increasing the supply of critical building materials by facilitating their production and distribution;

· Assisting the development and diffusion of appropriate building technologies that can be used by households at different levels of income.

The National Housing Policy has pointed at laws and regulations which inhibit housing construction. Among these are the Urban Land (Ceiling and Regulation) Act and the Rent Control Act. Uniformity of the legislation among the states would help achieve better administration, and might thus assist in increasing private-sector investments in housing. Furthermore, the Government must see to it that appropriate arrangements and measures are taken in order to achieve effective implementation of new legislation.

The National Housing Policy has a number of goals which are of importance for the poor (if implemented). The central and state governments will:

· Promote a more equal distribution of land in urban areas, and curb speculation in land and housing in consonance with macro-economic policies for efficient and equitable growth;

· Avoid forcible relocation or dishousing of slum dwellers;

· Encourage in situ upgrading, slum renovation and progressive housing development with conferment of occupancy or tenurial rights wherever feasible, and undertake selective relocation with community involvement only for clearance of priority sites in public interest;

· Expand the provision of water supply, sanitation and other basic services in slums and other settlements occupied by the poor;

· Ensure proper maintenance of amenities through community involvement and decentralized institutional arrangements;

· Promote incremental construction and upgrading by poorer households through access to land and services, through technical support, outlets for low-cost technology and materials, opportunities for skill-upgrading and access to housing finance on flexible terms;

· Provide night shelters and sanitary facilities for the footpath dwellers and the homeless;

· Encourage individuals and groups to construct houses for partial and full letting by access to land, institutional finance, enabling regulations and incentives in central, state and municipal taxation of property and incomes.

The emphasis of national delivery systems has shifted towards supply and management of land, rapid expansion of infrastructure, maintenance of housing stock, rental housing provision and special programmes for disadvantaged groups.

The National Commission on Urbanization has advocated a concentration of investments in settlements of various sizes which can generate economic momentum, sustain economic growth, promote balanced urbanization, expand housing activities and facilitate equitable provision of services. It is recognized that public and private investment in infrastructure to expand the supply of serviced land needs to be stepped up through enhanced budget provision as well as institutional finance. Organizational arrangements would be strengthened for mobilization and increasing the flow of funds for infrastructure.

Yet, all the above being said, according to Bhattacharya (1990), experts on housing in India see the policy statement of the National Housing Policy as a “paper tiger”, i.e., that it will have very little effect on the shelter situation of the low-income households. These experts believe that the approach should have been “bottom-up” instead of “top-down”. Moreover, housing should have been declared an industry and granted all the benefits that are given to industries, in order to boost housing construction and employment.

In the seventh Five Year Plan (1985-1990) it was stated that the responsibility for shelter delivery would gradually shift from the public to the private sector. It restricted the social housing activities to the Minimum Needs Programme for artisans and landless labourers only. The total Plan allocation for housing was barely 1.3 per cent of the budget. The eighth Plan approved 1.47 per cent. Targeting of the funds was sought by using the Minimum Needs Programme and a special programme for the scheduled castes and tribal populations. Critics of the present government policy, however, emphasize that it is wrong to privatize public agencies and to leave the poor and lower-income group at the mercy of the market forces (Bhattacharya, 1990).

A major concern of the authorities at central, state and local levels has for a long time been the growing urban slums, and the inadequate power available to public agencies to take over land and ameliorate the situation. The Slum Areas Act was passed in 1956 to rectify this situation. Several states have passed similar acts. These acts have been used for clearing public places such as pavements and congested business areas. The displaced populations were resettled in areas with minimum basic amenities, mostly at the outskirts of the cities. The emphasis of public policy shifted from slum clearance to slum upgrading in the early 1970s. Local authorities made in situ developments, and collected betterment charges from the beneficiaries. Yet, this policy could not be applied on a large scale for rehabilitation of slums, due to administrative difficulties and the long time required in acquiring land through the legal process. The capacity of public authorities to take possession of land by paying immediate compensation and then to launch shelter projects is rather limited, due to the inadequate financial and administrative support from the central and state governments. In Bombay for instance, only 23 slum pockets out of 800 were provided with improved facilities during this period. Moreover, the plots reserved for the poor often went to better-off households. Another reason for the limited success of this policy was that only the public sector could carry them out. The new shelter policy is based on such earlier failures. The record of implementing the recommendations of the GSS is not very impressive, but progress is being made in India.

4. Nigeria

In Nigeria, the provision of housing has generally been seen by policy-makers as something to be tolerated rather than desired. Housing was thus given low priority in development planning. Specific output targets have for instance always been set for agriculture, manufacturing, roads etc., but housing has been treated as a “social overhead” (Achunine, 1993). A review of the past housing policies and programmes of both the public and private sectors reveals that effective solutions to the shelter problems are yet to be found. It has been assumed in Nigeria that general economic growth would eventually solve these problems. The public sector has provided only about 10 per cent of the housing stock in the country. A new National Housing Policy was launched in 1991. The ideas included in this policy imply a redirection of past practices. Shelter was for instance transferred from the consumer to the regional development sector.

The ultimate goal of Nigeria's New National Housing Policy is to ensure that all Nigerians own or have access to decent housing, at affordable cost, by the year 2000. To achieve this laudable goal, the Government has decided to pursue the following policy objectives:

· Encourage and promote active participation in housing delivery by all tiers of government;

· Strengthen institutions within the system to render their operations more responsive to demand;

· Emphasize housing investments which satisfy basic needs;

· Encourage greater participation by the private sector in housing delivery.

The above objectives, among others, constitute the cardinal points for the implementation of the housing policy. To accomplish these objectives, the following strategies have been adopted:

· Establishment of an appropriate institutional framework to facilitate effective planning in housing supply;

· Restructuring all existing public institutions involved in housing delivery at the federal and state government levels with a view to making them more effective and responsive to the needs of citizens of the country;

· Revive existing laws and regulations such as the Land Use Decree, planning laws etc., to facilitate housing provision;

· Improve the finances and strengthen the executive capacity of local government to enable it to contribute more effectively in housing delivery;

· Mobilize private-sector participation in the provision of housing;

· Upgrade and rehabilitate low-quality or sub-standard houses in urban areas as a step towards improving the quality of the environment;

· Restructure the Federal Mortgage Bank of Nigeria to serve as an apex housing-finance institution;

· Mobilize savings through the establishment of a National Housing Fund;

· Ensure continuous flow of adequate funds from various sources into the apex institution for on-lending to other mortgage institutions;

· Encourage research into and promote the use of locally produced building materials as a means of reducing housing costs;

· Adoption of functional design standards to reduce costs and enhance socio-cultural acceptability, safety and security and privacy;

· Increase the number and improve the quality of the workforce and personnel needed in the housing sector;

· Utilize the location of housing estates and other residential neighbourhoods as an instrument for balanced population distribution in order to minimize associated problems of transport and services.

The Federal Government will initiate, define and coordinate the policy options and instruments for achieving the objectives in the housing sector, while the actual implementation will be undertaken by appropriate agencies at federal, state and local government levels. The Federal Government will formulate policy, coordinate, construct and monitor housing programmes and projects.

C. Housing needs

1. Mexico

The average annual population growth in Mexico is 2.1 per cent (1980-1991); the figures for urban areas is higher, 2.9 per cent. A total of 73 per cent of the country's population in 1991 was urban (World Bank, 1993b). The shelter deficit in Mexico as a whole, is estimated to 6 million units (1990). This includes families without shelter and those with inadequate housing conditions. Slum and squatter settlements have been estimated to constitute about half of the housing in urban Mexico (Connolly, 1990). Moreover, between 60 and 70 per cent of shelter construction is now made by the households themselves. In Mexico the modern sector employs less than a quarter of the labour force. This sector does not have the capacity to absorb the growing economically active population. Moreover, since 1982, the economic crisis in Mexico has led to a considerable reduction in industrial activity. In Mexico City, the percentage of the population living in self-built shacks increased from 2 in 1947 to 22 in 1952 and 50 in 1976. More than 10 million people may today live in slums in the city's metropolitan region. Between 30 and 40 per cent of all shelters are rented.

The annual need for new housing in Mexico in the 1990s is estimated to about 585,000, due to population growth (277,400) and deterioration of housing stock (308,000). During 1991, the public housing agencies provided altogether less than 348,000 housing units, i.e., only about 60 per cent of the estimated need. The number of houses built for the poor was 26,378.

To cover the needs of the poor, the government has set a construction goal of nearly 300,000 units a year. Yet, in the 1947-1990 period the public housing agencies made only about 804,000 units for the poor (see table 17). The need for shelter, particularly at the fringes of the big cities, is huge indeed. The poor, who typically make their dwellings by self-help, are confronted by several difficulties such as occupying land illegally; being in a weak bargaining position vis-is building materials traders; having insufficient technical assistance; and finding it hard to obtain credit. The settlements are often found in swamps, and on inhospitable salt-flats, steep hillsides and garbage dumps. The land and street layouts are chaotic and services are often lacking.

In 1982, it was estimated that the slum settlements in the metropolitan area of Mexico City contained between 3 and 3.5 million people out of a total population of about 14 million (Lombera, 1993). No reliable figures are available to update this estimate. The 1990 Census revealed that 575,865 people earned less than 1 minimum wage in the Federal District (a smaller area). It is not known how many of the about 2.5 million inhabitants in the rest of the metropolitan area earn below one minimum wage or live in slums. To obtain a figure comparable with the 1982 assessment, it is necessary to add those who earn more than one minimum wage and live in a slum. The number of squatters has undoubtedly increased, but legalization programmes have reduced the number of slum settlements. It is thus not given that the relative size of the slum population has increased. The metropolitan area's population is currently above 20 million.

About 22 per cent of the population of Mexico City earned less than one minimum wage in 1990 (a total of 63 per cent earned less than two minimum wages, see also table 7). Two minimum wages are thought to be the minimum required to afford a decent shelter. This is surprising, since estimates show that only 20 per cent of the settlements in the city are slums. The national average estimate, that 40 per cent of all dwelling units are situated in slums, may thus be closer to reality. As many as 64 per cent of the population in the metropolitan area is estimated to be residing in slums and squatter settlements (Schteingart, 1990). Regarding renting, it is likely that the national average of 27 per cent of the slum housing is much lower than the figure for big cities. In the city slums, 52 per cent of the people rent their accommodation. Furthermore, 23 per cent of all housing units have one or two rooms (45 per cent have three or less rooms). It may be concluded that the shelter needs of the poor are enormous, although exact figures are not available.

The city of Jalapa had a total of 285,758 inhabitants in 1991, on a total area of 3566 ha. The city has a considerable amount of vacant land for housing. Most of this land is private, and the subject of speculation. In the low-income settlements, land prices are still fairly low. It is estimated that 46 per cent of the population live in inferior, slum-like settlements: 45 per cent of the households in these settlements rent their accommodation. The housing problem is growing in Jalapa, with illegal slums being extended at the city fringes. The authorities have been unable to cope with the problem. It is expected that the changes in housing policy made will take time to reach the poor.

An estimated total of 31 per cent of the people living in the slums at the outskirts of Jalapa, earn one minimum wage or less (see also table 7). The land they occupy often poses severe technical limitations to the introduction of basic services. At the same time, existing water sources are polluted by industries and animal farms. The water distributed by road by the authorities or private traders is irregular. The muddy roads during the rainy season often make the water situation precarious. The introduction of sewerage systems for human excreta is possible in only 30 per cent of the slums. The extent of pollution of the groundwater from latrines is not known. Municipal solid-waste removal is seldom undertaken from these areas. Waste disposal is thus often made on vacant plots. A variety of illnesses are prevalent. The shelter needs of the poor are extensive.

Neither the public agencies, the private sector nor the squatters themselves have been able to solve the housing problems of the poor in Jalapa. The scarce public funds channelled to housing projects for the poor have often been used according to political affiliations. The available funding has been looked upon as business for enterprises with government contacts, rather than as projects to help the poor.

2. Indonesia

Indonesia's population grows at an average rate of 1.8 per cent per year while the urban population increase is about 5 per cent. The urban population constituted 31 per cent of the total in 1991 (UNDP, 1993). By the year 2000, it is projected that 85 million people will live in urban areas. The problems of urban growth are reaching an alarming stage. The shelter need is enormous and increasing in urban areas. It is estimated that just below 2 million dwelling units of all types are required annually in the next 20 years to meet the national housing need. A total of 750,000 units is required annually to meet the demand created by the population growth, while a further 100,000 units are required each year in the next 20 years, to alleviate the current housing shortage, estimated at 2 million units. The remainder is required to replace (350,000 units) or upgrade (700,000 units) the deteriorating housing stock.

The poor cannot afford ownership of a conventionally-made dwelling. Flats are not seen by them as an appropriate form of living. Even the government-built “very simple house” is beyond the reach of most of the poor. The poor with employment in the formal sector, e.g., the lowest paid government officials, can afford this kind of housing due to the fact that they can obtain government loans. The poor without access to formal credit are thus forced to look for a home in a kampung. The Government is not supporting the shelter need of the poor directly, only indirectly by improvements of the basic infrastructure such as footpaths, public toilets, garbage collection and water taps in the kampungs. In squatter areas no such improvements are made. Rental houses are the most popular shelter for the poor because of low cost and closeness to the location of work.

Table 11. Housing stock in Jakarta, by type (1990)

Housing type

Number of units














Source: Based on Herlianto, 1993.

Jakarta had a population of 8.2 millions in 1990, 12 per cent of which lived on the 6 per cent of the land characterized as slums. Estimates show that the city has an overall housing shortage of about 10 per cent (Yoewono, 1993). Other estimates indicate that as many as one fourth of all households are without decent shelter (see tables 11 and 12). Only 32 per cent of the dwellings had running water in 1985 (Gilbert and Gugler, 1992).

Table 12. Housing stock in Jakarta, by tenure (1990)

Tenure type

Number of units

















Source: Based on Herlianto, 1993.

Bandung is the third largest city in Indonesia, and capital of the West Java province. The city and its surrounding towns have grown into a metropolitan area with a population of just above 2 million in 1990 (in 1960 it was just below 1 million). About 23 per cent of the housing stock is located in slums (see table 13). The poor squat on riverbanks and along railway lines. The city's pavements and parks are also used for spending the night. So far, no special effort has been made for the poorest by the local authorities, except through a cooperation with Department of Social Affairs rehabilitation centres. This has had a very limited impact only on the shelter issue. The Government produces “simple houses” and “very simple houses” for the “poor.” These types of units are popular, but not affordable for the poor. The other main type of unit produced by the formal sector is flats. Yet, it has been shown that such housing reduces social interaction and limits the opportunities for using the living quarters for a shop or other home-based economic activities. And, as mentioned earlier, even subsidized flats are too expensive for the poor.

Table 13. Housing stock in Bandung, Indonesia (1990)

Housing type

Number of units














Source: Based on Herlianto, 1993.

3. India

The average annual growth of population in India is 2.1 per cent (1980-1991). The figure for urban areas is 3.7 per cent. A total of 27 per cent of India's population lived in urban areas in 1991 (World Bank, 1993b). The National Building Organisation estimated in 1981 that 30 million of urban dwellers lived in slums. This was nearly 19 per cent of the urban population in that year (Kundu, 1993). It is estimated that the urban shortage of shelter is between 6 and 13 million units (Ribeiro, 1993). The Seventh National Plan projected the urban housing shortage to be 5.9 million units. The Birla Institute, however, estimated the shortage to be 22.1 million units as early as in 1981. If, instead of using this norm-based calculation, a demand-based approach was adopted, the housing shortage would be low. This is due to the low priority people can place on using their scarce resources for shelter improvement purposes.

The major problem of housing in urban India is not the absence of any kind of structure, but its poor quality and the non-availability of basic services. More than one third of the urban households, for instance, have no access to a latrine, and 18 per cent have access to a service latrine only. Furthermore, 45 per cent have only one room, while less than 1 per cent of all urban households are categorized in Indian statistics as houseless (Kundu, 1993). Households with some kind of temporary structures have been excluded from the houseless category. This includes many pavement dwellers. There has, however, been an increase in the number of substandard urban dwellings. Kundu (1993) finds that 30 per cent of the urban poor live in dilapidated shack-like structures and mud dwellings (see table 14). The total production of houses through State budgetary support over the last 30 years in India is a tiny proportion of the country's total shelter construction, and a small fraction of the real requirement of shelter for the poor (Sundaram, 1986). Yet, public-sector building of formal housing is not the best way of spending the scarce government funds available for shelter delivery. The approach should concentrate on identifying strategies to involve the poor in a sustainable manner. A shift in attitude among urban planners is necessary. Furthermore, economic statistics regarding housing include only investments in formal housing built within the approved shelter process. This implies that the economic value of low-income households' own construction and shelter improvements are unrecognized. This lowers the overall contribution of the housing sector to the economy in official data relative to what is the case in actual fact. In this context it should be remembered that government provision in India accounts for a tiny fraction of the total housing activity only - less than 5 per cent (Pugh, 1990).

The housing problem in India is not one of housing in the conventional sense, but of shelter and of avoiding extreme congestion. Although shelter is not the poor people's first priority, it should be an indispensable component of any development strategy. In India's big cities amenities are everywhere grossly inadequate. The housing shortage is so severe that a solution lies far into the future. In the biggest cities, 65 per cent of all households live in one room or less. Private house-building of “affordable shelter” is the only realistic way forward (Rajeswari and others, 1987). The Government should thus not attempt to do what people have already demonstrated that they can and will do themselves (Dhar, 1988).

Table 14. Slum population of various cities in India, 1981 census


Total population

Slum population







Greater Bombay
































Source: Bhattacharya, 1990.

The share of slum population in the total urban population has been shown to be positively correlated with the size of towns and cities in India (Kaul, 1991). The situation is probably similar for several other developing countries. In India the proportion of the population living in slums in towns below 100,000 inhabitants is about 10 per cent. The figures for towns between 100,000 and 1 million inhabitants and the cities of 1 million and more are 20 and almost 33 per cent respectively. The poor undoubtedly perceive their chances of survival to be better in the mega-cities.

In sum, despite considerable investment and efforts over successive Plan periods in India, the housing problem continues to be daunting in terms of the large number of homeless households, the rapid growth of slums and unauthorized colonies, spiralling prices and rents for land and houses, rampant speculation, deficient availability of water, sanitation and basic services to the bulk of the population, and the increasing struggle of the poor and vulnerable people to secure affordable and adequate shelter. The rapid urban population growth has led to worsening congestion and overcrowding in small houses, steady growth of slums and squatter settlements and severe pressure on public services. This is aggravated by institutional deficiencies of housing agencies and local authorities, and insufficient attention to the shelter needs of the poor.

Bombay is the financial capital of India. Modern business buildings dominate the central city landscape. At the same time, every patch of undeveloped land is filled with makeshift shelters. In 1991, Bombay was the largest urban area in India with a population of 12.6 millions. Before independence the population growth rate was manageable in terms of basic services. High-rise dwellings (of 4-6 floors, called chawls) were constructed. These buildings have since deteriorated into slums (see box 18). After independence, the growth of economic activities and population accelerated. A recent decline in the growth rate is probably due to state-level strategies to redirect growth to other cities. The Maharashtra Housing and Area Development Authority estimated in 1992 that about 25 per cent of the households in Bombay are poor (Ribeiro 1993), and that 37 per cent of the population live in slums. Many of the older buildings in Bombay are now obsolete. Overcrowding and over-use have made them become shaky structures with bad flooring, peeled-off walls and wobbling staircases. Leaky toilets and water-taps are a common sight. The environment has become very unhygienic. Many buildings are beyond repair, and several of them collapse each year during the wet monsoon. The Rent Act of 1947 is partly to blame for this state of affairs. Since tenants cannot be evicted, and rent increases cannot be made according to inflation rates, owners reduce maintenance costs and investors turn away from low-cost house-building.

Box 18. Creation of a slum

“But looking at some instances it would appear that the people are also to blame. Take the example of Kherwadi, a veritable slum in West Bombay suburbs. Quite some efforts were made to settle a group of tanner families there and give them a good plot per family, a fine town planning scheme, and so on. But instead of taking advantage of these and the facilities for improving their tanning trade with modern facilities offered by the Tanning Institute located in this place, they turned into landlords, built small dingy sheds, rented them out at very high rents and themselves lived and still live in the most deplorable conditions. In a place where a hundred households were to be settled there are now about a thousand households of which about a hundred are landlord households and nine hundred are tenant households. Having initially accepted the terms and conditions of a sub-human life there, the tenants now clamour for better facilities.

... in fairness to the government, that in trying to use its powers and develop its policies it has to take into account the low income of the people, the huge gap between income and shelter costs, high land costs, land speculation, lack of individual and national finance and savings, absence of a well organized building industry, backwardness in architecture and the use of local materials, public pressures, political influences, influences in the selection of personnel and consequent inefficient administration, corruption, etc.”

Source: Desai and Pillai, 1990.

In 1991, Bombay had a housing need of more than 2.5 million dwelling units. More than three quarters of the population live in one-room dwellings. Pugh (1990) estimates that 45,000 families join the squatter settlements every year. The dwellings in the slums are more frequently contracted out than self-built. The creation of this kind of houses - constructed of waste materials - is a large industry, the value of which is not included in GDP figures. The houses are affordable, but insanitary and a conspicuous sign of poverty. Moreover, the number of pavement dwellers is increasing. Some of these people are second- and even third-generation pavement dwellers. There are over 200,000 households living in the streets and 430,000 household in rundown walk-ups or in squatter settlements. Today the public sector cannot cope with the problem of indecent and unhealthy shelter conditions. Collaboration between the Government, private entrepreneurs, NGOs, CBOs and the poor is thus necessary.

Box 19. Bombay: a people's manifesto for housing

“ A.

1. Stop all evictions;
2. Regularize and legalize all slums;
3. Strictly implement the Urban Land (Ceiling and Regulations) Act;
4. Nationalise ownership of all land.


1. Implement 'Sites and Services' scheme for housing economically weaker sections;
2. Nationalise production, marketing and distribution of building materials;
3. Subsidise costs of building materials used for housing economically weaker sections.


1. All basic services must be provided and subsequently maintained by the state. These services include provision of water, electricity, toilets, sewage, drainage and garbage clearance;

2. Amenities like schools, hospitals, ration shops, creches and play grounds should be provided and maintained by the state for economically weaker sections.


1. People must have access to all information kept with the government, must have the right to participate in all forums and have the right to decide all issues that affect them;

2. Priority to women and the aged in housing;

3. Census must cover all and ration cards must be given to all;

4. Rejection of cut-off dates for recognition of slums.

... But the story of struggle is endless. In December 1986, hutments below Napean Sea fly-over were demolished to make way for the Priyadarshini Park. After Nivara Hakk's intervention, and after the struggle that ensued, the displaced people were taken to Chembur with the promise of an alternative plot. But they were dumped on the pavement there. Their struggle is still on. Sangharsh Nagar, a small slum colony at Mankhurd, had half-a-dozen demolitions in the last six months. They have forcibly occupied a strip of swamp land and are determined to fight it out to secure the right to live and exist. The struggle goes on.”

Source: Das and Gonsalves, 1987.

Interestingly, there has been a drop in migration to Bombay during the last years. Urban population growth is now mostly caused by natural population growth. Furthermore, there were in 1985 about 50,000 acres of vacant land. Das and Gonsalves (1987) thus conclude that since even the government of Maharashtra admits that Bombay can accommodate up to 20 million people without congested housing schemes, the real problem is not too many people or too high migration or too little land, but the unequal distribution of land. A people's manifesto (see box 19) has been worked out to suggest reforms in city planning, conducive of poor people's interests.

The houseless include people who are unemployed, street children, widows, destitute, old and handicapped people but also self-employed street vendors. It is estimated that there are at least 10,000 street children. The houseless can be found in most parts of the city (see box 20). They concentrate close to public toilets and water taps. The shelters they make of rags and plastic waste are often demolished by the authorities to clear the streets. Community life on the pavement is normally organized. People of the same caste and religion stay together. They hardly interact with neighbours living in regular houses. They move around very little but stay close to their source of livelihood. A part of the houseless is made up of itinerant worker-families following available construction employment. Not all pavement dwellers are poor but belong to the lower-income group with very low or no ability to save. There are many NGOs working with and for the houseless in Bombay. The activities of the Don Bosco shelter project (box 21) and the Youth for Unity and Voluntary Action (see box 22) exemplify this.

Box 20. Pavement dwelling in Bombay

“ 'Would you even spit here?' Triveni asks in disgust. She lives in a covered cot on one of the pavements in Kamathipura, the red-light district in central Bombay. She gestures to the narrow space between her abode and the next cot, which is occupied by a man. Both covered beds jut out at right angles to the street.

'I have to do everything here,' Triveni continues angrily. 'I cook here and at the crack of dawn I bathe here in full view of everybody. And when the rains come, the water collects and rises right up to our cots. What are we supposed to do?'

Triveni is one of the people who are known in India as 'pavement dwellers.' They are truly 'the wretched of the city,' existing a notch lower down the social scale than even the despised slum dwellers. The people of the pavements evoke intense feelings among Bombay's better-off citizens - revulsion, hostility, sometimes pity.'

Drought is often a reason for migration to Bombay. The state of Maharashtra is drought-prone. People are then pushed off their small land holding and cannot find work on other farms. Their only alternative to starvation is to move to towns and cities, put up a tent of rags on a pavement and look for ways to earn something to eat.

The Supreme Court in India ruled in 1985 that people had no fundamental right to set up a shelter in the roadside. They could however only be evicted on certain conditions such as being given notice and regarding certain cases without being given alternative site to live on.

In 1986, 300 pavement dwelling families were forcibly moved to Dindoshi, a northern suburb of Bombay. The municipality provided them with 15m2 plots. Later another 1,500 families were moved there. Many women complained - 'All they want to do is to get rid of us. But if they do, who will sweep and swab their floors and wash their dirty dishes and clothes for such a pittance?'“

Source: D'Monte, 1989.

Box 21. The Don Bosco shelter project

The Don Bosco shelter project provides free accommodation for 40 boys aged 10 to 21 years. The boys work in the waste-recycling trade. They have to pay Rs. 2.50 a day for food ($US 0.10). The project enables them to sleep, cook and to play.

Calcutta, which for decades was the largest urban area in India, now has fewer people than Bombay. The population is 11 million. The squatter settlements are largely found at the outskirts of the city on marginal land. Upgrading through popular participation has had limited success. The pavement dwellers in the city have likewise received little assistance but much compassion. It is through the bustee improvements the city has made valuable experiences. A bustee is a rental low-rise slum.

Box 22. Youth for Unity and Voluntary Action (YUVA)

The Youth for Unity and Voluntary Action (YUVA) is an NGO that works with the youth and women on the streets. It attempts to mobilize the houseless to organize themselves and to fight for improved living conditions. YUVA has successfully trained youths as urban animators. The five-month training consists of various issues related to development activities, problems of the environment and organization. They are also trained in drama and puppet shows.

Kundu (1993) estimates that 45 per cent of the households in Calcutta are poor, whereas Pugh (1990) states that 35 per cent of the population is below the official Indian poverty line. About 68 per cent of the households live in one-room accommodation. The shelter need is above 2 million units. In addition, the existing housing stock is deteriorating. Public and private house construction are both insufficient. In this light, the improvement and extension of the bustees are important. The bustees cover only about 7 per cent of the land area of the city, but house almost a third of the population. The bustee settlements are not only places of residence, but, as slums elsewhere, they constitute centres of informal economic activities. The slums in Calcutta have a three-tier tenancy structure. The owner of the land leases plots to people who make a dwelling. The dwelling is then rented by a poor family. This complex system of property rights has impeded historical attempts to reform the bustees. Some bustee settlements have stable populations with supportive social networks, although drunkenness, crime and prostitution may be rampant. Other bustees contain recently arrived migrants. These settlements tend to consist of hutments of split bamboos, with mud floor and roofs of makeshift materials. They occupy low-lying sites prone to monsoon flooding. Bucket latrines shared by up to 20 families are typical in such unimproved bustees. Those who cannot afford even this kind of shelter are forced to reside on pavements. In fact, pavement dwelling has become a way of life for many people in Calcutta.

In 1991, Delhi had a population of more than 8 million, compared with 1.4 million in 1951. The rate of growth is high, in spite of efforts to curb urban migration by limiting employment opportunities in the formal sector. It is estimated that by 2001 there will be 14 million people in the city with 6 million more in townships within the wider metropolitan area. Most of the land in and around the city is under public ownership and control. The Delhi Development Authority has a near monopoly on land allocation and housing development. There is a shortage of land in the city. At the fringes land is sold illegally, and poor people squat in several spontaneous settlements without basic services. In the inner-city, pavement dwelling is widespread. There is a need for close to 2 million shelter units. Of these, 130,000 are needed by pavement dwellers and 600,000 by other squatters in the more central parts of the city. There are few options for renting but in the slums at the outskirts of the city. It is estimated that the poor constitute 25 per cent and the lower-income group 35 per cent of the population. About 49 per cent of the population live in one-room accommodation (Pugh, 1990), and at least 70 per cent of the households cannot afford the concrete houses built by the private sector.

Due to the large urban migration (more than 200,000 annually), the rents in the slums are increasing rapidly even for single-room tenements. Even this simple accommodation is now un-affordable to many migrants. The number of houseless people is thus growing. Night-shelters are made for these people. This accommodation provides toilets, blankets and jute mattresses at a subsidized rate of Rs. 2 per night. The shelters are open from 7 p.m. to 8 a.m. At present only 3462 people can be accommodated in such shelters in the city as a whole. There are, however, plans to extend this type of shelter for individuals to more places and also to cater for women. In the 1992-1997 period, 10 additional buildings will be erected, paid for by the Central Government.

Despite various attempts by the Government over the years, the urban sanitation situation in Delhi has recorded little improvement. An estimated 67 per cent of the population is without a sewerage connection, while 47 per cent is without a latrine. Those most affected by lack of facilities are squatters and pavement dwellers. The manual scavenging of “night-soil” is recognized as a social evil. In 1985 there were 500,000 bucket toilets in the city. Resource constraints, however, rule out the extension of the sewage system to the entire population. A low-cost sanitation programme has therefore been initiated.

4. Nigeria

The most typical dwelling type in Nigeria is the “rooming” type. By this is meant renting (or owning) a room in a house (not a flat). About 65 per cent of all households live in this type of dwelling. A further 15 per cent live in flats while only 17 per cent of the households live in “separate dwellings” (Achunine, 1993). In the urban areas only one third of the households own their accommodation (see table 15). Given an urban population growth rate of 5 per cent, the annual urban requirement for housing units in the 1990s is nearly 400,000 units. In addition, there is an enormous need for upgrading and replacement of existing housing stock. Achunine (1993) estimates that in Nigeria as a whole there is an annual need of 1.4 million hosing units.

Table 15. Housing tenure, Nigeria, 1985 (percentage)















Source: Achunine, 1993.

Table 16. Estimated income groups, Nigeria (1992)

Income group

Share of households

Monthly income












2,500 +

$US 1 = Naira 30

Source: Based on Achunine, 1993.

The 30 per cent of the population with the lowest incomes do not have sufficient funds to exercise an effective demand in the formal housing market. This group of people is probably in basic-needs poverty (see table 16). The 20 per cent lowest on the income scale earn below N50 per month. This group will be poor in a narrow basic-needs sense. The impact of economic conditions in the 1980s have moreover forced many of Nigeria's more well-off families down into lower income categories. The purchasing power of a typical middle-income family was reduced by a factor of 8 during the 1980s (Achunine, 1993).

The most common form of tenure is short-term lease. Leases of less than five years account for 56 per cent in two low-income settlements in Lagos studied by Aina (1990). This form of tenure is unusual in customary Yoruba land tenure practice. It represents a specific response of the urban poor to the cost of and availability of land. Extensive subletting of land exists with the owners of the land living elsewhere without any direct contact with the tenants. Improvements of infrastructure and shelter are thus impeded. The occupiers of the land do not see it as worthwhile to build concrete, solid and, thus, expensive structures on the rented land. And the owners of the land are not interested in investing in the land but are waiting for the value of the land to increase. They hold the land for speculative reasons. In short, there is an enormous need in this and other urban areas of Nigeria for land which can be distributed and to which low-income groups can obtain a title.

In no city in Nigeria is the housing failure more manifest than in Lagos. There the traditional forms of dwellings are not acceptable. The legal acquisition of land is far beyond the means of most people, not only the low-income groups. Opportunities of employment are very limited, and the population is growing at an alarming rate. Prospective renters have to queue for more than a year on waiting lists before they can be considered for a rental unit. And rents are extremely high, the average worker has to spend as much as 40 per cent of his/her income on rent. To aspire to own a house is a dream realized only by the top 5 per cent of the income groups. Often a down-payment has to be provided two years in advance for a flat.

It is estimated that in the 1976-1985 period nearly 850,000 new housing units were required in the Lagos metropolitan area. Only 82,000 units were actually built. An estimated 100,000 new people arrive in Lagos each year. This may be compared with building plans for 1991 of 4800 housing units. It is no surprise then that three fourths of all families live in one room only, as one survey estimates (Achunine, 1993).

Over the last decades the solution attempted regarding the housing problem in Lagos has mainly been in the form of rent controls, often through military edicts. This has failed to achieve the desired purpose. Rent controls collapse in the face of an extreme housing shortage, and supply and demand are then left to decide the prices.

D. Shelter delivery

1. Mexico

In Mexico, as elsewhere, most housing programmes favour the better-off people. Yet, the number of loans for low-cost housing has increased during the 1980s, from 2166 in 1980 to 48,225 in 1985 and 104,070 in 1991. These loans are directed at the poor and the lower-income group, both in sites-and-services and upgrading schemes. The production of houses in the public programmes has grown slowly and the bulk of public low-income housing has been constructed during the last 10 years (see tables 17 and 18). Most of the dwellings made are for private ownership. The building of rental housing is seen as non-profitable. Housing-finance agencies did not contribute to the construction of rental houses in 1989 and 1990. The poorest are thus very much left to manage on their own resources.

Table 17. Low-income public housing production in Mexico


Annual number of low-income housing units constructed

Percentage of all completed units

























Note: The figures (not the total) are annual averages for the periods shown. “Low-income” refers to 65 or 60 per cent of the population with the lowest incomes. The source is not clear at this point.

Source: Based on Carmona, 1993.

Many squatter settlements (often called “invasions” in Mexico) in Mexico City have been recognized de facto and have received some basic services. Lawful property rights have, however, not been given to the residents, but recognition of a settlement has been granted to community leaders affiliated to the political party in power (Connolly, 1990). After 1970, regularization was integrated into the urban policy agenda. A number of state-level regularization agencies and housing institutes now exist. Land has been taken over, and state governments are the major controlling force for regularization and the supply of new land from expropriated community properties. The residents' committees elected every four years at block, neighbourhood and district levels have consultative powers vis-is the authorities. It is the neighbourhood-level committee that is allowed to negotiate on matters of regularization and services to the exclusion of other popularly formed community organizations. Authorities are now in charge of selling serviced as well as unserviced plots to individuals. Still, clandestine developers manage to operate.

Table 18. Public-sector investment in housing, Mexico


Investment as percentage of GNP











Source: Carmona, 1993.

Box 23. Valle de Chalco: self-help construction

In Valle de Chalco, on both sides of the highway that connects Mexico City with Puebla, illegal settlers formed 19 colonies from 1970 to 1980. In 1990 the total population of the area was 252,413. The first occupation involved the acquisition of ejido land by coyotes. Later the peasants started to sell land. Legalization of occupied land was a constant problem for the squatters. It is, for instance, a rule that an area in which the development of services is impossible, cannot be legalized. Another problem for the settlers is the fact that the area is divided between the Federal District and the State of Mexico.

In 58 per cent of the families, no member has regular employment. Of those having employment, 43 per cent earn one minimum wage or less. No organization represents the interests of the settlers. Even mutual aid is said to be a rare phenomenon in the area. All of the dwellings have been made through self-help, in some cases with the help of paid workers for certain parts of the construction. The settlers have managed, individually, to obtain illegal access to land and to make a shelter. Yet, they have not solved the problem of water provision, drainage and paved streets.

Box 24. Santo Domingo de los Reyes: making money at low-income levels

The project involves 4000 squatters. Since rental housing in Mexico City is scarce and expensive, the option available to the poor is to settle on a vacant plot in the outskirts of the city. Such a plot has no basic services. The squatter must pay an illegal organization for the right to use the plot, or obtain it from a local-level leader specializing in illegal land occupation. The dwelling is then constructed - little by little - through self-help. The leader will also organize the squatters to make certain infrastructure for themselves such as footpaths, drainage and water supply. Often such leaders are professional invaders, called coyotes, who make their living by invading lands. The squatters must pay the leaders for “protection.” Community leaders have an important task in reducing the influence and power of coyotes. They also function as intermediaries between the poor on the one hand and the coyotes and public authorities on the other, as reported from Quito, Ecuador (Burgwal, 1993).

In Santo Domingo de los Reyes the squatters made 90 to 200 m2 plots and streets and footpaths, without maps or an overall plan. Posts for lighting were erected, and electricity was obtained illegally. A water supply was organized by truck delivery. In the beginning water was carried from far away. Schools, shops and market places were also built. The squatters were tightly organized by the leaders. Later, the settlers themselves were able to take over the negotiations with the authorities, and to obtain advances in the process of legalization. At that time they tried to get rid of the leaders. They have not been completely successful in this, but they have managed, to some extent, to curb the power of the leaders.

The legalization process moves slowly and is so far only achieved for parts of the settlement. Due to the development of the area, the price for the plots has increased, thus benefiting the owners. The compensation the authorities now have to pay for the land is at such a level that the squatters cannot afford it without large subsidies being given. Many plots are thus taken over by lower-income households and even by middle-income people.

Squatter settlements without an outside person or group fixing the “invasion” were typical in the past. Today new squatter settlements are normally arranged by so-called coyotes or pirates on communal or ejido lands (see boxes 23 and 24). The ejido land is communal holdings distributed under land-reform programmes in Mexico. The peasants have use-right over such land. Urban growth has made much of this land close to cities more valuable as building sites than as agricultural land. It then becomes profitable to allow settlements of poor people on the land. The ejido land holders will claim compensation from the settlers. The settlers may later have to pay the Government for the land. Speculators involved in these illegal subdivisions of land are able to avoid the cost of service provision. The settlers move in without basic services, and then ask the authorities to provide them. Tenure will remain insecure, and the authorities may evict the settlers if that is politically possible. Many settlements which are in breach of planning laws, develop due to the profit motive of speculators and to politicians seeking popularity and votes. Some squatter communities are able to organize themselves and to build rudimentary basic infrastructure and services gradually. This helps to establish the settlement's legitimacy, although not in a legal sense. Other settlements remain in a state of extreme deprivation, lacking all facilities, and hence the poor find themselves under a permanent threat of eviction.

Box 25. Jalapa: upgrading at the urban fringe

This project in Jalapa is an integrated development of four squatter settlements' shelter, education and health facilities and productive and cultural activities. The settlements lack basic services and infrastructure and the housing conditions are bad. 83 per cent of the family heads earn one minimum wage or less. The inhabitants are organized, and have improved the settlements to some degree. No public programme is directed to these areas.

The Union of Settlers, Tenants and Solicitants of Housing of the State of Veracruz and CENVI have been engaged in the four settlements. They promote participative planning by mobilizing the inhabitants. Technical, financial and administrative support is given to the collective efforts of shelter and infrastructure improvements. The collaboration between CBOs, the NGO and a public institution is promising. The successful projects in Jalapa were possible due to support from CENVI.

2. Indonesia

In Indonesia, the National Urban Housing Development Corporation (Perum-Perumnas) implements housing projects aimed at the poor and lower-income group. In line with the State National Development Goals of 1988, shelter should be affordable without neglecting the minimum standard of decent housing. Units of 21 m2 are the most needed in Perum-Perumnas low-rise housing. For flat housing, 36 m2 units are the most popular. The public sector construction targets have, however, not been met. In the planning period of Repelita-IV (1984-1989) 140,000 units should have been provided by Perum-Perumnas. The actual figure was about half that. The number of house ownership credits certified by the State Savings Bank (BTN) going to Perum-Perumnas has declined during the 1980s - from 46 per cent at the beginning to 20 per cent at the end of the decade. The Government has gradually moved towards building more high-rise flat housing. This is, however, an unpopular form of shelter due not only to cost but also for socio-cultural reasons. Yet, even the cost of the cheapest housing unit supported by the formal housing sector in Indonesia, the “very simple house,” amounts to at least Rupiahs 3 million (see table 19). This amounts to 12 times the annual income below which a head of a household is categorized as being poor. Even the interests of such an investment is beyond the reach of the poor. It is thus not possible for the poor in Indonesia to afford even the cheapest formal-sector housing unit. Yet, while concluding the above, it should be kept in mind that the average household may have several sources of income, not only the one of the head of the household, which is used in the statistics on income groups.

A special programme for the poor is the Serviced Plots Program. In Repelita-V, a total of 50,000 units are planned. The credit from BTN for a 21-m2 core house requires a 10-per cent down-payment and it carries a 12-per cent annual interest over 20 years. The beneficiaries may build their shelter gradually by own-effort. Public amenities such as footpaths and toilets are provided. Furthermore, the plan is to make 20,000 simple rental dwellings for those unable to obtain credit. The rental houses and flats will be subsidized by the Government.

Yet, even this scheme is, by definition, not affordable for the poor. Even if they can afford a down-payment (of at least two annual incomes), the annual interests on this loan would amount to at least Rp. 540,000, while the annual income defining the poor is about Rp. 250,000.

Table 19. Average formal-sector house prices, Indonesia (1992)

Size (m2)

Constructed by

House type



(Rp. million)

Private companies
(Rp. million)

Very simple



































Source: Herlianto, 1993.

The Government initiated KIP to upgrade the environment in which the poor live. This began during Repelita-I in Jakarta, with World Bank support. It was later extended to cover 10 other cities, 200 towns and about 1000 small towns. Although KIP was not aiming at solving the housing shortage per se, it has shown that improvements of infrastructure and services in the kampungs encourage people to improve their dwellings. In general, KIP resulted in an increase in the number of households selling food of different kinds from their dwelling place (Batarfie, 1987). This provides important additional income, although the amounts may be small. A very important aspect here is the extension of the dwellings, which people undertake, by adding a room for letting purposes. Improved direct support for such extensions should be adopted by governments in developing countries (UNCHS, forthcoming). This would help the poorest families who cannot afford more than to rent a small room. The KIP is also important in a more indirect sense, because when the Government has improved a kampung, it is regarded as a legal settlement. People are then eligible for government subsidies. These programmes in Indonesia are well conceived but far below the needed magnitude for sheltering the poor. The Government's ability to provide housing for the poor and lower-income groups is very limited. At the same time as there is a fast growing need for more subsidized shelter, the public sector's capability to cope is decreasing. It is thus essential to stimulate the traditional spirit of mutual aid in small communities.

3. India

In India, shelter delivery for the poor is very difficult. The requirement of cost recovery, the rising cost of building materials and the decline in government funds available for housing have led to the adoption of shelter solutions that price the poor out of the delivery system. In general, housing in urban areas in India is primarily the responsibility of individual families, about 90 per cent of the investments and 70 per cent of the supply of housing units are made by the households themselves (Kundu, 1993).

Moreover, access to plots is controlled by mafia-style promoters. In 1981 between 20 and 26 per cent of the urban population lived in slums and squatter settlements. The Government earlier tried to house the poor in multi-storeyed flats built on slum land or to resettle them in sites-and-services areas. Both failed, due among other things to cost and location. The thrust has recently been to improve the infrastructure of the slums in order to remove the worst environmental deficiencies. The dwelling situation is thus not improving. Squatting is made more difficult on non-used public and private land except at the urban fringe. Since the distance from the outskirts of the cities to places of economic opportunities in the centre is often large and transport costly and time-consuming, many among the poor are forced to stay on the pavements. Many lower-income group people may also be found in the illegal and unregulated settlements at the urban fringes. It seems essential that a process of accelerated resettlement of pavement dwellers is initiated by public authorities in the big cities in India. Resettlement success can only be achieved through a constructive dialogue with the pavement communities. It seems that the location of resettlement is the fundamental issue. The first priority of most squatters is to live near the source of livelihood (see box 26). Bahri (1988) for instance, arrived at this conclusion from a household survey in Hyderabad.

According to Pugh (1990), squatters in Delhi, who had been resettled, found themselves far away from job opportunities. The utility services were scarcely better than in their original locations. More importantly, family incomes tended to decrease due to increased difficulties for women to find work in the neighbourhood and to travel expenses incurred by men in searching for work in and around the city centre. As much as half of those who become resettled, may sell their plot rights illegally to higher-income households. They then revert to squatter settlements closer to their places of work opportunities.

During the 1980s there were several programmes of slum upgrading in the cities, benefiting the poor with simple but temporary shelter (see boxes 27 and 28). The combined size of these programmes is not possible to assess, nor is the percentage of beneficiaries of the poor urban population. This is due to lack of adequate and comparable data. In general, there was a rapid growth of housing stock during the 1970s and 1980s, simultaneously with a relatively declining public-sector investment in housing. The private sector is thus responsible for a growing share of shelter provision in India. Yet, small contractors are not operating on a wide scale in India. Their quality is often low, and people have little confidence in them. People are thus generally forced to use public agencies or large builders for house construction. This is expensive, and the poor are left to provide themselves with shelter. In the 1980s there was a rapid increase in the number of shacks in urban areas (a much faster growth than in the 1970s). About 20 per cent of the housing units produced during this period belong to the formal housing market. Sixteen per cent of all urban households live in mud dwellings, while 26 per cent live in temporary dwellings.

Box 26. A redevelopment programme in Dharavi, Bombay

The Prime Minister's Great Project is a programme initiated by the Government of Maharashtra in 1985 to improve existing slums, particularly Dharavi, the biggest slum in Asia. Dharavi covers an area of about 330 acres of low-lying, marshy land. Tanneries have grown in this area adjoining an old and small fishing village. Earlier single-storeyed semi-permanent labour shelters were constructed there. Then huts were put up in every vacant spot. The huts are made of old tin, bamboos etc. There is an extreme overcrowding and appalling sanitary conditions. Legalization and better shelter conditions are sought in this programme. Infrastructure and shelter improvement loans will be provided to the poor. The population of Dharavi is estimated to be about 250,000. A part of it is on private land. Those who wanted to take part in the programme, had to organize themselves in CBOs. The accommodation provided consisted of one-room tenements with a toilet. Each tenement costs around Rs. 60,000. A subsidy of Rs. 5400 is given. An interest free loan of Rs. 7100 and a loan of Rs. 20,000 at low interest are provided. The rest of the needed cash must be arrange by the people themselves. This is very difficult for the poor. This kind of high-rise flat housing programme although subsidized, is of limited use to the poor.

At present, the purchase of urban land for housing by the poor is out of the question. Only the public housing agencies can provide land to the poor through, for instance, cross-subsidizes. Recently, land rights have been given to slum dwellers by some local authorities, often for political reasons. Those getting a plot can then obtain housing loans. For the squatters occupying private land, this option is not available. It is evident that the public sector has been unable to reach but a small segment of the urban poor.

Box 27. Ekta Vihar: squatter upgrading in New Delhi

The experience of limited success in several resettlement projects in Delhi made the Government try upgrading existing slum settlements on the site when the land is not required for other urban purposes. The Ekta Vihar project has the objective of promoting self-help and community involvement through technical and financial assistance. Social and economic activity support are included as well.

Some residents of Ekta Vihar asked an NGO, Action for Securing Health for All (ASHA), to open a clinic in their slum. This led to the formation of a CBO and contact with EIUS. EIUS provided the settlement with hand pumps and community toilets. ASHA started to train community health workers. The squatters then wanted to improve the infrastructure in the area and their shelters. The haphazard arrangement of dwellings and the narrow footpaths were a hindrance to upgrading. ASHA organized the community into a registered CBO with elected leaders. ASHA also acted as an intermediary between the CBO and public institutions. This was necessary because the squatters had no confidence in the government policies for upgrading. They had been evicted several times from different areas.

The authorities were persuaded to allow the 475 households to reside permanently on the land, and land rights were transferred to EIUS. The allocation of plots was made jointly by EIUS, ASHA and the CBO. The residents chose their neighbours. The standard on basic services and the dwellings were kept low to discourage selling. EIUS together with ASHA made three demonstration dwellings to show the construction technique and building materials. A loan of Rs. 5000 at 4 per cent annual interest was given to each household. The development of the land was given to the residents as a grant. The CBO was entrusted with the control of any resale of plots. Every household has to pay Rs. 30 a month for environmental upkeep of the community. The low default on this charge and the repayment of the loans are due to the fact that the residents see the positive transformation of their habitat. It is also important to recognize that the area has a fairly central location with relatively easy access to the residents' established sources of livelihood.

Box 28. The Bustee Improvement Programme (BIP), Calcutta

The Bustee Improvement Programme (BIP) became operational in 1972. A century of neglect had left much to be done in the bustees. By 1986, the BIP had improved about two thirds of Calcutta's slums substantially, benefiting around 3 million people. The aim of the Programme is to improve the environment, especially the sanitation situation, without influencing the tenancy system. This system is formal, i.e., it is within the legal and tax framework. The improvements sought are: to construct latrines with a septic tank for every 25 persons or on the basis of a cluster of dwellings; to provide a water tap for every 100 persons with at least 20 litres per capita and day; to make drainage ditches connected to underground sewerage systems; to pave footpaths and minor roads; and to provide garbage collection and bath facilities at convenient locations. By 1987, a total of 460,000 slum dwellers had got better basic services under this programme.

Bustee organizations typically consist of the inhabitants of several clusters of houses. These organizations are approached by the authorities in order to identify the slum dwellers' specific needs in that particular area, and to secure their cooperation in, for instance, rearrangement of the dwellings around the water tap and toilet facilities provided. Included in this model of slum upgrading is a health-care programme. To enable the poorest to take part in the programme, it was realized that income-generating opportunities had to be initiated for them. A small-scale enterprise programme was therefore initiated to find viable service and production activities that could be started with credit assistance and technical advice.

The Programme's multifaceted approach has improved the living conditions of the poor even though the tenurial right to plots has not been given. The cost of the Programme is to be recovered by a tax on the beneficiaries. Yet, the collection of this tax has not been pushed. The maintenance of the improvements made is thus a problem, since the authorities are unwilling to take on this responsibility as long as the recovery rate of betterment charges is so low. Yet, it must be said that BIP has had some degree of success. It is however insufficient, and lacks several necessary dimensions (Moitra and Samajdar, 1987). The Programme should be harmonized with the regional economic development programme; the complex land-tenure situation should be solved; the rising rentals forcing the poor to find cheaper accommodation should be checked; and the balance between cost recovery for replicability and necessary subsidies for the poorest must be found.

The soundness of the fundamental idea behind the BIP should, however, be stressed. The public expenditures on BIP have reached the poor. The typical experience in site-and-service and resettlement programmes of land and shelter being to a large extent traded up from the poor to higher-income groups have largely been avoided. Moreover, redevelopment projects often reduce the number of people that can be sheltered on a piece of land. The extremely high congestion in the bustees makes it, in some sense an efficient use of land. The most important thing to note, however, is that the number of dwellings in Calcutta probably would have declined, if redevelopment had been initiated instead of the BIP.

The “night-shelters” in Delhi are meeting a demand for a cheap place for individuals to spend the night. Cost recovery may be difficult to achieve in this kind of accommodation. Replicability is thus limited. A chain of night-shelters, which also caters to women, could for instance be made and operated by NGOs. The need for a place to sleep close to business opportunities is widespread among the poor. In India there has been a steady decline in the supply of rental accommodation, mostly due to the Rent Control Act. The rental options for the poor are fraught with exploitation by “slum lords” and other intermediaries. In the formal sector there has been little construction of tenements for rent since rental rates were frozen for long periods, reducing profitability in this kind of housing investment.

4. Nigeria

The Land Use Decree has been a major hindrance for shelter provision in Nigeria. It has not released land speedily and adequately enough to satisfy the basic land needs for housing construction. Moreover, its cumbersome procedures are intimidating for the poor, for whom formal access to land is very difficult. A prime example of a law that is quite obsolete is the Nigerian Town and Country Planning Ordinance of 1946. This law controls the major part of the process of translating unused land into land for shelter construction. The main criticism of the law is that it does not offer sufficient participation by relevant actors; it does not reflect modern land-use planning concepts, usages and practices; and it has not been synchronized with the provision of the Land Use Decree of 1978.

Between 50 and 60 per cent of total house-construction costs in Nigeria are those for building materials. Currently, much of the materials is imported. Local capacity and capability of building-material production must be developed. The reasons for the high building-material costs, besides importation, are: high demand owing to the oil boom; adoption of high building standards by the more well-off population; over-priced contracts; and inefficient distribution systems, aggravated by intermediaries. The Government is moving away from direct production of building materials and intends to limit its role to facilitating private sector production. Furthermore, the Government will encourage simple, imaginative and functional building design, and realistic specifications and space standards to facilitate cost reductions, affordability and acceptability; and will develop appropriate technology and support the use of local building materials.

It is obvious that housing supply in Nigeria has bee unable to keep up with housing demand. There are a number of reasons for this situation, among which are:

· Unstreamlined and incompetent mortgage arrangements;

· High cost of land and sometimes utter inaccessibility to it;

· High construction costs occasioned by the continuous upward rise in the cost of building materials;

· Lack of skilled labour;

· Inadequate housing finance;

· Too much attention is paid to prestigious projects. Most public servants in Nigeria have, for example, built prestigious mansions in their villages that are usually occupied for not more than three months each year. These are regarded as status symbols while economically, they constitute relatively dormant investments;

· Most builders of residential accommodation are set on recouping their invested capital within the shortest possible time;

· In enforcing development control measures, very high standards are usually set which have a dampening effect on housing provision. Appropriate building standards can do much to create a safe and pleasant environment. Yet, misconceived, they can contribute to depressing the living standards of the poor. If such regulations are inappropriate and set standards too high for existing income levels, their primary effect will be to reduce the amount of housing that is available at prices the people can afford.

Other factors that adversely affect the low-income groups include:

· High rates of inflation;

· Inadequate infrastructural facilities.

· Limited access to serviced land and the difficulties of obtaining title (certificate of occupancy);

· Lack of finance.

The lack of finance is the most important factor inhibiting the access of low-income Nigerians to decent housing, through high interest rates; unaffordable down payments; low earning power, making it impossible for low-income groups to afford the monthly repayment rates; and inability to provide guarantors acceptable to the mortgage institutions. Consequently, only a few urban dwellers - and even fewer rural dwellers - have so far benefited from mortgage loans by the commercial banks, the Federal Mortgage Bank of Nigeria and the state housing corporations.

E. Actors and programmes

1. Mexico

Although self-help shelter programmes started in Mexico in the 1960s, their importance was not evident until the late 1970s, and then mainly through the sites-and-service projects developed by Fondo Nacional de Habitaciones Populares (FONHAPO), and directed at the needs of the low-income groups by supporting societies and cooperatives. There are three sources of supply of housing for the poor and low-income group - the public, private business and household sectors. Private business may belong both to the formal (typically large-scale construction firms) and the informal (small unregistered enterprises, but usually independent craftspeople) sectors. Most low-cost urban housing is made by the households themselves, with the assistance of craftspeople for certain more difficult construction tasks. The Government assists the middle- and lower-income groups - as well as the poor with housing - through FONHAPO (see, for example, box 30). It provides technical and financial assistance to its target group which is almost 80 per cent of the country's population (see also section II.F.1). FONHAPO works through legally constituted organizations. Every family is, however, responsible for its loans directly to the institution.

Box 29. PROFOPEC (Programa de Fraccionamiento Popular de Ecatepec): a project involving coordination at the highest level

The project involved 8100 families from different parts of Mexico City. In 1987 poor families started to squat on private land at the outskirts of the city, in an area called Zona V de Ecatepec. Other groups soon joined the illegal occupation. The number passed 15,000 families. Families from nearby areas lacking basic amenities also settled there. The area is of bad quality, being the floor of a dried-out lake. The challenge of the project was to reconcile the low quality and limited quantity of land with the massive demand.

Of the squatters, 53 per cent had an income below 1.5 minimum wages, and a further 36 per cent between 1.5 and 2.4 minimum wages. Representatives of the squatters negotiated with SEDUE and the Government. The Government expropriated a nearby plot of almost 156 ha. and handed it over to PROFOPEC. The president of PROFOPEC was the Governor of the State of Mexico and board members were the Secretary of SEDUE and the President of the Municipality of Ecatepec. Their ability to “cut corners” resolved the complex legal aspects. Different NGOs assisted later in various ways of securing services and shelter provision.

The beneficiaries paid only between 5 per cent and 10 per cent of the cost as down-payment according to house type. In the “core house” programme, the beneficiaries had to pay 38 per cent of the family head's income in monthly payment. The figure in the upgrading programme was 25 per cent. These arrangements were agreed upon by the beneficiaries' organizations according to the squatters' economic capabilities and the prices of the different shelter types. FONHAPO assisted these organizations in obtaining the down-payments by making each family pay a weekly sum to accounts political parties had opened for them. The rivalry of the political parties was reflected also on this issue of shelter provision for the poor. Each week the parties published figures for what they had managed to induce the squatters to pay.

The project legalized the tenancy of the plot holders and provided and improved drainage and water supply. PROFOPEC also made 12 wells. The number of “core houses” constructed was 6227. The houses were either of 18 m2 or 36 m2. The walls were of cement and the roof of prefabricated slabs. This project responded to the ideas of the beneficiaries, and was in accordance with the popular housing ideal in Mexico of a private one-family shelter.

Box 30. El Arenal: a housing project for single mothers and young couples

This project in Mexico City covers about 1000 families from the inner city. The background of the project was the awareness of overcrowding revealed by the 1985 earthquakes. After the earthquakes the families lived in tents and then in emergency housing units. These units were blocking traffic and hampering trade. The families were therefore resettled near the international airport on land bought by the Government. The Reconstruction Programme of Popular Housing organized and developed the area. The dwellings were 24 m2 each. The construction and materials of the dwellings were simple and cheap. In 1987 the inhabitants decided to upgrade their shelter with more solid materials. There were water taps in the streets and footpaths.

The neighbours were at first organized in groups according to footpaths. Later a larger association was formed but not legally registered. 700 families joined. The association managed to obtain several special arrangements with FONHAPO and other government institutions. The loans from FONHAPO were granted to individuals but administered collectively by the neighbourhood organizations. The monthly payments were fixed at 32 per cent of the income of each family. The mortgage was extended to 13 years. The land was used as collateral for the loans. Building materials were obtained at low prices through a programme carried out by the Secretaria de Desarrollo Urbano y Ecologia (SEDUE), in collaboration with private enterprises.

Box 31. Renovaciabitacional Popular (RHP): a rehabilitation programme

The official number of families rendered homeless by the earthquake in Mexico City in 1985 is 100,000. Most of the houses destroyed were old rented tenements with rents frozen for a long time. The maintenance of the buildings was bad, and their resistance to the earthquake thus relatively low.

The people who lost their accommodation were able to organize themselves and put pressure on the Government for action. The city authorities expropriated about 1000 properties in the affected areas. These properties became the basis of the RHP programme. The beneficiaries were only the tenants of expropriated properties. The affordability of the beneficiaries was taken into account. Only in special cases were extra subsidies granted. Building norms, in particular density norms, were flexible in order to accommodate the original amount of residents. The programme was administered plot by plot, thus securing the community feeling of the population. Interestingly, the RHP programme was very different from the rest of the emergency housing construction. The positive experiences with the RHP programme made it a model for subsequent housing projects, such as Casa Propia (see box 32).

Box 32. Casa Propia: tenants and owners

The beneficiaries of this project in Mexico City are 10,000 families living in flats and rooms in apartment buildings. They rent their accommodation. The rents were not allowed to rise. These two facts led to lack of maintenance of the buildings. This lowered the buildings' resistance to earthquakes. The buildings were not included in the official reconstruction programme after the 1985 earthquakes.

The project assists the rehabilitation of the flats. To be eligible for support, the tenants must have a legal document proving the tenancy, and a unit must not exceed 69.7 m2. Furthermore, the tenants must not earn more than 2.5 minimum wages. The down-payment is 10 per cent of the total cost, the repayment period is seven years and the monthly payment is 10 per cent of a beneficiary's income. The project has shown that improvements and housing growth are possible in inner-city areas.

The urban renewal project in inner Mexico City developed after the 1985 earthquakes was successful due to political and financial support. Popular participation was important in this project, and it can be a model for socially-oriented shelter delivery for the urban lower-income group.

The National Solidarity Programme in Mexico consists of a limited support to the improvement of housing through the provision of cheap building materials to the poor and lower-income group (see box 29).

The private-sector contractors show no interest in housing provision for the poor because of the low profitability and high risk of default in payments. CBOs have only appeared recently in Mexican urban shelter delivery. The 1985 earthquakes in Mexico City led to the formation and later popularization of CBOs in inner-city renewal (see boxes 31 and 32). Cooperative housing is not found among the poor (see box 33. In sum, most dwellings are built by the people themselves with assistance from relatives and neighbours.

The NGOs in Mexico have provided valuable assistance regarding shelter to the poor. This has been possible because NGOs can effectively make the linkage between public institutions' regulations and resources and people's social needs. They can supply technical know-how and train local-level leaders and people generally. They do not, however, help with financial resources for house-building.

2. Indonesia

The private sector in Indonesia, both large- and small-scale, does not cater to the poor's shelter needs. It is the Government through the Perum-Perumnas that assists the poor. Yet, it is able to address the needs of only a percentage of the poor. In fact, less than 15 per cent of urban housing is made by formal-sector actors. The major part of the shelter provided, stems from the people themselves, sometimes with the assistance of CBOs and NGOs. This, unfortunately, implies that the security of land tenure is very low, which again results in a low motivation for house improvement.

Box 33. Calpulli del Valle

The cooperative in Calpulli del Valle in Mexico City was joined by 900 families from various parts of the city. Nearly half of the families (46 per cent) earn less than 1.5 minimum wages, and a further 30 per cent earn between 1.5 and 2 minimum wages. The NGO COPEVI (Centro Operacional de Vivienda y Poblamiento) gave technical and legal advice to the cooperative, which then acquired an area of 52 ha in Coalcalco. According to land use plans, 19 ha of this land was suitable for housing. The local authority of Coalcalco decided that no funds were available for developing of the land. FONHAPO, however, does not require public development of lands before providing loans. COPEVI thus worked out an alternative way of developing the land. The solution included wells and tanks for drainage. After six years of negotiations and paperwork with federal, state and municipality authorities the cooperative obtained permission to make the wells.

The construction of the wells was paid by the cooperative together with an NGO called FUSCOVI (Fomento Solidario de la Vivienda). FONHAPO then granted a loan for the land acquisition. The permit for the tanks for drainage is however not yet given by the authorities. They require a secondary treatment for the liquid waste. Such a treatment facility is unaffordable for the cooperative. No core houses have thus been constructed so far.

In Indonesia, beneficiaries are not permitted to use their plot in regulated areas for other than residence (UNCHS, 1991c). For the poor, such a rule (if enforced) would keep many from taking advantage of shelter projects.

An important factor of enabling shelter provision among the poor, besides mutual aid in building, is “arisan”, a collective lottery for housing purposes. Once a week the members of an arisan pay money or the equivalent in rice to, usually, a woman coordinator. A weekly draw decides which person will receive that week's collection. The system is quite widespread among the poor in urban areas also (see box 41).

The Community Based Low-cost Housing programme is created to mobilize communities to improve their shelter and the environment. Support is also given to stimulate social and economic activities. The target group of the programme is the people evicted from squatter areas. The programme particularly helps in forming CBOs, and in facilitating these organizations' access to plots and credit under the programmes of Perum-Perumnas. The programme is limited to two smaller areas in Jakarta and Bandung (see boxes 34 and 35).

Box 34. Tamansari: private rental housing

The location of this low-rise rental housing in Bandung is strategic. It is near to transport, public offices, educational facilities and markets. In 1990, the Tamansari area was chosen by municipal authorities in collaboration with the Institute of Human Settlements of the Ministry of Public Works to be given “rental house improvement credit.” The project was carried out together with an existing CBO. Every household could obtain the equivalent of Rp. 550,000 over a period of three years - without interest - in building materials. The purpose was to construct additional rooms to ordinary families' dwelling. The room(s) should then be let to poor people. A profile of the tenants shows that they are street vendors, casual workers, low-level government employees and students. The occupation density of the houses is very high. The houses are of low quality. The rental tariffs are from Rp. 16,500 to Rp. 30,000 per month ($US 8-15). The monthly incomes of the tenants vary from Rp. 60,000 to Rp. 150,000, i.e., they are classified as poor or lower-income households.

Cooperative housing has been very limited in Indonesia. Government assistance has popularized cooperative practices to some degree such as the workers' cooperatives. Yet, these cooperatives do not cover the poor, but only the lower-income group.

The KIP in Indonesia has been successful in improving the infrastructure and basic services in several urban slum areas. It may be sensible to have such a programme which does not include plot regulation and shelter construction. This is so because the severe complexity in giving squatters a legal title to the land they occupy would indefinitely postpone an upgrading of the physical and social infrastructure. If such land development was carried out by employing the poor in the community with labour-intensive methods, they would gain both by some temporary income and an improved environment. In addition, there would be a mobilizing effect regarding organization and mutual aid. The impact may also be a stimulation of individual shelter repair and extension.

Box 35. Bina Karya: a housing foundation

The Bina Karya housing foundation, an NGO in Bandung, aims at providing houses for fired textile workers. The former workers rented their accommodation. Due to their irregular and low incomes, the 120 people were not entitled to credit from BTN. With money given by the factory and with the assistance of the foundation the workers started a shelter project. Bina Karya started as a consumer cooperative. With the help of LPSM (a community self-help institute) it soon developed into a housing cooperative. Land was acquired at the fringe of Bandung, and infrastructure and shelter made gradually according to the financial and work ability of the residents. The dwelling units are 42 m2, on a plot of 96 m2. Space has been made available for sport facilities, a mosque and a multipurpose community building. Financial support has been received from a foreign NGO.

A major lesson from this experience is that, in order to attract the attention of an NGO, it is essential for community shelter projects to collaborate with a national institute or organization.

KIP's improvement of infrastructure in Jakarta resulted in plot holders increasing their spending on their dwellings. It also led, as could be expected, to rising land prices and higher rentals. This, unfortunately, has had a negative effect for the poor, many of whom are renting their dwellings. No positive change was reported in the residents' incomes (Taylor, 1987). The infrastructure and services in slum areas are, however, now deteriorating. KIP cannot keep pace with the growing slum populations. The Urban Renewal Project, with flat housing, is thus being implemented in the slums. Yet, the poor can seldom afford flats, and they do prefer low-rise housing. Still, the so-called Klender flat housing is a solution favoured by the Government to overcome the shelter problem of the poor (see boxes 36 and 37).

3. India

The Central Government in India has played a major role in providing shelter and basic services to the urban poor, although in the Indian federal system the states have the responsibility of implementing housing and urban policies.

Box 36. Flat housing in Jakarta: Kebon Kacang

A total of eight four-floor buildings with 600 dwelling units was built. Among these units, 368 were without separate bedrooms. The city government, together with the Perum-Perumnas, has targeted this project at poor street vendors and hawkers. There is a cross-subsidy involved, as some of the units are offered at a higher price. A hire-purchase project is operating for those who can afford it ($US 30-35 monthly instalment over 20 years). There are several similar flat housing projects in Jakarta.

Box 37. Klender flat housing

This is the largest flat housing project in Jakarta, situated in Jatinegara. It is surrounded by kampungs. The aims of the project are to maximize land use, maintain a healthy environment with open green belts, and be affordable to the poor and lower-income group. Perum-Perumnas assists in securing loans for the families from BTN. Evidence shows that only the lower-income group can meet the expenses, although these are kept low. The expenses are higher than rented accommodation in the neighbouring kampungs.

The Ministry of Urban Development is the apex policy-making body for urban development, including housing activities. It overviews the state-level programmes, lays down policy perspectives and builds up the data systems for the country, besides launching projects for the benefit of certain targeted groups.

Public sector housing is basically the responsibility of the state governments. The housing boards are the implementing bodies. In Maharashtra there is, in addition, the City Industrial Development Corporation (CIDCO) which has housing as one of its major activities. A fixed percentage of CIDCO's housing is set aside for low-income groups. The cities' development authorities take the major responsibility for designing and executing public housing programmes.

The role of the primary cooperative societies in providing housing has been increasing over the years, particularly in the large cities. Cooperatives get preferential treatment in obtaining land and finance from public-sector agencies. Due to the scarcity of land and capital, it has become difficult to acquire land and finance through individual efforts. Now cooperatives can also be found in the slum areas of the largest cities.

The large-scale private contractors mostly cater to the higher income groups. Small-scale contractors operate as individuals or firms. They take the responsibility of designing a house and getting the plan approved by the local authorities. A license is not required to operate in the low-income construction market. The absence of proper registration, accountability through formal contracts and a legal system for resolving conflicts speedily have got in the way of a healthy growth of a small contracting system. Still, the small contractors are essential, because they allow for incremental house building. Normally, poor households must build their house over an extended period of time.

The involvement of NGOs in the shelter sector has been of varying intensity. At the national level, there are few NGOs working on shelter issues whose objective is to benefit the poor. The National Campaign for Housing Rights is an exception. This NGO seeks the recognition of housing as a basic human right. At the grass-roots level, there are many NGOs in India facilitating upgrading of slums, mobilizing and organizing the poor for savings, credit and common action in general.

The Environment Improvement of Urban Slums (EIUS) programme has been in operation for a long time in India. At present, EIUS is a major programme of the Central Government designed only for the physical improvement of slums. The goals are to provide a water tap for every 150 people; to open drains for the outflow of waste-water; to build a community bath for every 50 people; to erect a lavatory seat for every 50 people; to widen and pave existing lanes; and to put up poles for street lighting 30 metres apart. The per capita expenditure under the programme was raised in 1985 to Rs. 300. In the Seventh Five Year Plan, slums in which two thirds of the families earned less than Rs. 250 per month, were selected for improvements. Slums on private land are also, in some cases, included in the programme. Security of tenure for the plots is given to the slum dwellers (on public lands only) parallel with the basic infrastructure improvements. The residents are then responsible for maintaining the assets provided to them. Under the Slum Clearance and improvement Programme, slum dwellers were earlier provided with an alternative plot at sites away from the existing slums. The problems were, however, lack of sufficient government funds and public land. Upgrading was then favoured under the so-called Minimum Needs Programme launched in 1972 and later transferred to the states. The programme consisted of physical improvements of the slums, i.e., water taps, drains, latrines, footpaths and sometimes also street lighting. The expenditure per inhabitant in this programme has unfortunately been low throughout.

Not only is a large part of the urban population in India deprived of acceptable housing, an equally large part has remained without access to safe drinking water and sanitation facilities. Many urban development programmes were therefore designed in the 1980s with the provision of water supply and sanitation as their major components. The term “shelter” implying a plot of land with a temporary roof and certain basic services such as water supply and sanitation was coined during that period. It was, however, recognized that a poor country such as India cannot, even under the best of circumstances, afford the provision of shelter and basic services to its vast majority of poor people, for the time being at least.

UNICEF has assisted in urban slum environment betterment. Three UNICEF-assisted programmes were in 1985 brought under a single programme called the Urban Basic Services. The programme works through CBOs in the slums, and seeks to improve health and nutrition; water supply and sanitation; education of children and training of community volunteers; and community participation in local associations. The focus of the programme is on women and children. Again the maintenance of the facilities provided is entirely left to the users. The Central Government, state authorities and UNICEF finance the programme on a 20:40:40 ratio. The programme has a community approach, involving the participation of the beneficiaries.

The Urban Community Development Programme (UCD) has the objective of involving municipalities in the provision of basic facilities in urban slums. The Programme's use of community participation in maintenance reduces public expenditures, so freeing funds for further improvements elsewhere. At present, the Central Government and UNICEF are assisting the Programme only marginally. Municipalities are thus getting funds for UCD from their own resources and through the state governments and foreign agencies.

Low-cost housing is also a component under several other urban community development projects. A plot title is given to the beneficiaries, a layout-plan of a neighbourhood is made, and dwellings constructed. Yet, only poor squatters on government land can be assisted, due to the need for tenure rights. The Manila Milan Women's Cooperative in India has shown that poor households can save. The amounts saved, however, are far from the required sum to acquire land and build a house. The savings can be used as bridge finance during emergencies and for small shelter repairs only. Governments should encourage and support the establishment of local so-called barefoot banks by NGOs and CBOs.

A low-cost sanitation programme, with technical cooperation from UNDP, is involved in erecting simple latrines. This programme has been adopted in many states. At present, the programme focuses on eliminating manual scavenging. States and local authorities also undertake projects on their own, providing shelter, water supply and sanitation facilities to the poor through upgrading and sites-and-services (see box 38). Besides Indian finance, the World Bank is an important source of funds at these levels.

In sites-and-services projects construction costs may be partially provided for (or not at all). The combination of efforts by the people, various voluntary organizations and government institutions is the hallmark of these projects. Many families, however, cannot afford to build a proper dwelling. They are forced to live in makeshift arrangements on their plots for a long time, or to sell their plot.

Programmes, such as the Slum Improvement Programme (SIP) and the Slum Upgradation Programme (SUP), are specifically directed at the poor. These programmes are supported by the World Bank. One aspect of the programmes is to relate the level of services and standard of dwellings to the affordability of the beneficiaries. This is, however, in many cases a sufficient input to raise the value of the land, thereby forcing a displacement of the original residents. The SIP involves merely physical improvement with a standard package of basic amenities. Cost recovery from the beneficiaries is now an essential part of these projects. Since the slums must be compatible with the official land-use restrictions, about half the slums in the big cities are disqualified for upgrading. In the SUP, plot titles are given to individual families. In some cases land is also leased to communities. A home improvement loan may also be obtained under this programme. Unlike the SIP, there are no direct subsidies under SUP.

Box 38. Sulabh Shauchalayas in Delhi

This programme aims to convert existing bucket latrines into two-pit water-sealed pour-flush public latrines: to construct baths and urinals on a pay-and-use basis; to utilize human excreta to generate bio-gas; and to rehabilitate scavengers by (raining them in various vocations. The programme will maintain the community latrines and baths for a 30-year period. The facilities are serviced around the clock The users pay Rs 0.50 for use. Women, the handicapped and the destitute have free access. The authorities only contribute the water supply and electricity. The maintenance of the facilities in poor areas is paid by a cross-subsidy from the money collected from the use of similar facilities in better-off areas. This programme shows that an NGO can provide an essential basic service with minimal government expenditure. Moreover it is evident that the poor are prepared to pay for a service. The continued training of masons and plumbers undertaken by the NGO, as well as a high rate of cost recovery ensures the replicability of the programme.

From the above, it is clear that in India there are several options in shelter delivery for the urban poor. As in Mexico and Indonesia, the major investments in the construction of dwellings in India are made by the people themselves. According to US AID (1989), 70 per cent of urban housing supply is being met by the private formal and informal sectors. Cooperative housing associations are not able to assist the poor. Their domain remains restricted to the higher- and middle-income groups, with a few minor exceptions in slums in Ahmedabad, Bombay and Delhi. Small-scale contractors are the only actors that work for the poor and lower-income groups, because their shelter construction normally is incremental, i.e., carried out over time according to available funds.

In several Indian cities, slum dwellers have organized themselves in CBOs. Some grants, loans and technical support have become available to these organizations under various programmes. The Government has often facilitated assistance from NGOs to the CBOs. CBOs are typically formed for the purpose of protest against possible eviction or for fighting a legal battle. Many CBOs have also been formed in the 1980s, in response to a requirement inherent in slum improvement programmes. CBOs have a tendency to consist of people belonging to the same background. In this way, a fairly strong community bond may exist. Yet, it should be remembered that community participation tends to be informal, i.e., not formalized into a legal body. The role of local-level leaders and NGOs is critical in accomplishing sustained community participation by establishing legally recognized CBOs. In the years to come the role of CBOs will be even more important, due to the government decision of a gradual withdrawal from direct shelter provision (see box 39).

Punervaas is a shelter movement in India which brings together local authorities, financial institutions and NGOs, with the aim of supporting the poor in the cities to help themselves in improving their living conditions (Kaul, 1991). Punervaas is a catalyst in the process of forming multi-purpose cooperatives in small slum communities. The movement also aims at creating attitudes of thrift and arrangements for credit, in order for small savings to be mobilized. Loans for small-scale enterprises will then become available to the poor. An innovative and utmost important task undertaken by the movement is to familiarize NGOs with specific slum communities. In this way the NGOs can give more appropriate assistance more efficiently.

Box 39. SPARC: relocation of pavement dwellers in Bombay

SPARC attempts to empower women to make and achieve their own alternatives to pavement dwelling and to design their homes and even to plan entire settlements. The aim is to place shelter developments under the control of people themselves, through new patterns of collective leadership in the communities. This is seen to be essential to change the fact that in nearly every shelter programme, more than half of the original beneficiaries sell their plot and dwelling. SPARC is of the opinion that if women play a major role in shelter programmes, this will not happen to the same degree. Today men are the target and decision-makers of housing programmes. Since women are the ones who make a shelter into a home and also more and more often work out the family's package of survival strategies, they should have a larger say in the planning of settlements. The situation for women is, however, the opposite - they cannot get housing finance or obtain a plot title even when they have a greater stake in gaining a permanent dwelling.

The city and state authorities maintain that there is no land available for resettling pavement dwellers. The alternative left for them was thus to help themselves. SPARC meant that this could be achieved if they understood clearly what was required, in terms of resources, policies and procedures, and then negotiated with the authorities. Each pavement community had to be organized with a collective leadership to avoid internal exploitation and “divide and rule” policies by external forces. With the help of SPARC, land was found at the outskirts of the city. Yet, the people found it to be expensive and too far from their sources of livelihood. The latter point was especially affecting women. The majority of the beneficiaries had no opportunity to examine the physical site beforehand and could not calculate the economic impact of the move. The result was that many could not repay the loans, and ended up selling their newly-acquired plot and returning to their earlier place of living.

After this experience SPARC initiated a training process for women in groups of 15-25. They were taken to different vacant lands in the city, and the criteria for suitability of the areas were discussed. The important outcome of this process was that people should resettle as whole communities, and that training in building skills were necessary. Since shelter provision for the poor must depend on self-help in construction, it is essential for quality housing that both men and women acquire skills through outside assistance in training. Moreover, in the decisions on the location of new settlements for the poor their work pattern, especially women's, should be considered. Space for income-generating activities in the home is also of significance for many women.

After an assessment of the various housing programmes for the poor and lower-income group in India Kundu (1993) concludes: “... success in terms of their targets has been extremely limited, although certain states have done better than others....” The National Commission on Urbanisation (NCU) has made a similar observation. NCU (1988) states that even the main targets were missed. It furthermore concluded that there was a high degree of inflexibility in the programmes and schemes.

The liberalization in the urban sector - including relaxation of the administrative and legislative controls - weakens public control over land, and thus restricts the capacity of public agencies to find land for sheltering the poor.

4. Nigeria

In Nigeria, there are only very few large-scale contractors engaged in the provision of housing. There is, however, a large number of small contractors that have contributed considerably to the shelter-delivery process. They have built for the middle-and low-income groups. NGOs, CBOs and cooperatives have no tradition in shelter provision in Nigeria, and are even today nearly totally absent. The public sector has so far concentrated its effort on owner-occupied housing, and thus on the better-off section of the population. The criticism of public housing has been: high unit costs; slow rate of output; low quality of production; inadequate attention to site location; and much “red tape” in administration. Achunine (1993) concludes that the few government housing schemes implemented, have not been of benefit to the poor or to the lower-income group.

Experience shows that the provision of sufficient infrastructure and basic services in Nigeria are beyond the financial capability of local and national authorities. Private developers have thus started to develop large tracts of land for housing construction. Funding of infrastructure now depends more and more on capital markets. Payments for infrastructure will therefore have to be collected primarily from the users. There is, however, an in-built cross-subsidization of land on some publicly developed land at the urban fringes. Other types of subsidies are uncommon in the housing sector in Nigeria.

F. Financing shelter

1. Mexico

The scale of poverty in Mexico makes it impossible for few of the urban poor to save anything at all. FONHAPO has a credit scheme for housing the poor, “Credit to the word”. Normally, this is given with the support of an NGO and combined with a self-help project. The programme has been successful in communities having a building capability. It is, however, mainly a rural programme.

In addition to FONHAPO, FICAPRO (Fideicomiso Casa Propria), is the only public finance agency serving the poor, i.e., those with an income below one minimum wage. Its target group are those living in the Mexico City metropolitan area who rent their shelter and who have an income of less than 2.5 minimum wages. Yet, these agencies' share of total housing finance in Mexico is rather low, FONHAPO being the larger of the two, distributing 3.9 per cent of all housing finance in Mexico in 1991. FONHAPO was, however, responsible for 12.9 per cent of all completed housing units in 1991.

Box 40. Los Mil Millones: quick solutions for poor groups

This project provided housing for 1404 families in Mexico City with incomes of less than 1.5 minimum wages. These families were not organized and they lived in various places. They were working as peasants, casual labourers, street vendors, domestic workers or they were unemployed. They had applied for shelter support to CAPROMOR (Casa Propria para Morelenses). CAPROMOR acquired land from the State land reserve. This land was then used as guarantee to obtain a loan from FONHAPO. CAPROMOR furthermore helped the families to establish a legal housing association which took over the responsibility for the loans which were in the name of the beneficiaries. The families earning less than 0.5 minimum wage received a subsidy from CAPROMOR. Another agency provided certain essential building materials. A contribution that really facilitated success was the delivery of “construction material packages” containing the necessary materials for the self-help construction of a “core house” including also basic electric, sanitary and water installations.

The largest housing-finance agencies, INFONAVIT (Instituto Nacional de Fondo de Vivienda para los Trabajadores) and FOVI (Fondo de Operaci Descuento Bancario para la Vivienda), which between them were responsible for 72.2 per cent of all housing loans and 59.5 per cent of all completed housing units in Mexico in 1991, do not cater for the poor.

Yet, only 25 per cent of FONHAPO's and 12 per cent of FICAPRO's loans are given to applicants with an income below one minimum wage. An additional 60 and 74 per cent respectively are given to applicants earning between one and two minimum wages. Although the Mexican Government avoids giving direct subsidies to housing, the low recovery rate of FONHAPO loans implies a government subsidy.

2. Indonesia

Housing finance to the poor and lower-income groups is in extremely short supply in Indonesia. Major reasons for this are high inflation and increasing land prices. The lack of low-interest credit for housing is recognized as a major impediment to the provision of urban shelter. In the present situation of structural adjustment policies limiting the funds and space for government action, few options for a change can be seen. It is thus a serious problem that the poor have such a low ability to save for housing purposes. Many among the poor cannot afford to save at all. Another group may save a little with a food security aim. In Indonesia those among the poor who may have a sufficient income to save for shelter improvement, are often reluctant to place their money in a bank. This is a fact in spite of the Government's effort to popularize a savings habit by several special programmes through BTN.

The housing loan system catering to the poor is directed at house ownership in regulated areas with basic services. Since these areas are at the fringes of the cities, many poor people, who otherwise could afford such housing, cannot take these loans because transport to workplaces in the city centre is too problematic.

Housing subsidies are given to people with low monthly incomes to enable them to acquire a very simple dwelling on a serviced plot. Most of the poor do not qualify for this subsidy because they either have no employer, have no savings account or because their incomes are irregular. They have to depend on the limited assistance provided by the arisan system (see box 41).

For the urban poor in Indonesia, lack of money is a dominant barrier in their struggle for decent shelter. The low-cost houses built by Perum-Perumnas are still beyond the reach of the poorest groups. They are only able to afford rental housing in the kampungs or to erect a makeshift shack in a squatter area. The Government provides a subsidy to families that live in kampungs and want to extend their dwellings with the view to providing others with rented accommodation. However, the reduced ability of the Government to subsidize house-building seriously affects the poor. The fundamental problem is the inadequate land regulation, unclear titles, lack of tenure and the operations of speculators and intermediaries. This is an obstacle to expanded investments in housing. Since the informal sector and the people themselves are the main force in shelter provision, the granting of land-tenure rights to urban poor households - especially in kampung areas - should be enforced. The problem is that granting building permits to squatters is regarded as legal approval of their occupation of the land. No real improvements of poor people's shelter are thus possible without regularizing land, and this is extremely difficult in practice. Even the housing provided by Perum-Perumnas and supported by BTN is today beyond the reach of the poor, due to a substantial reduction of the implicit subsidy. Many lower-income households have benefited by this government effort at shelter provision. This relieves some of the downward demand pressure on land and houses for rent for the poor.

Box 41. Palasari kampung, Bandung: arisan housing

A total of 35 arisan groups were formed in Palasari to provide shelter. Each member paid Rp. 4000 ($US 2) per week. The requirements to become a member of an arisan group are permanent residence, bad or no house, and irregular income. Each week there is a draw, which leaves the money to one person in the group, except for a small part that goes to the coordinator and to community purposes, such as improvements of footpaths, drainage and garbage collection.

3. India

Experience has shown that some of the poor in India do have the ability to save. Investments in housing may, however, not be their first or even second priority when it conies to making use of the savings. In general, the share of household savings used for the creation of physical assets has gone down over the years. The direct investment by households in house construction has declined.

Yet, the majority of the poor do not save, and have no ability to do so. In a sample survey in Bombay, 80 per cent of the poor with a shelter and 92 per cent of those without, did not report any savings at all (Acharya and Trikha, 1978). A study by the Bombay Metropolitan. Region Development Authority (1981) similarly concluded that for “industrial labour” expenditure exceeded income, leading to perpetual indebtedness. Ahmad (1982) found that households in Delhi earning Rs. 1400 per month or less were negative savers, while those earning between Rs. 1400 and 2428 were non-savers. The percentages of people belonging to these categories have been estimated to be 33 and 25 per cent, respectively. The conclusion is that the thesis of the possibility to mobilize savings among the poor for investments in housing by designing appropriate schemes and institutions is questionable. Moreover, according to an analysis made by Ribeiro (1993), projects requiring the urban poor to spend more than 11 per cent of their income on housing cannot be sustained given their present expenditure pattern.

The idea that the savings potential among the poor can be effectively mobilized through NGOs, cooperative societies and other institutional arrangements has been put forward in international literature, according to Kundu (1993). It is also thought that the urban poor have the capacity to make substantial down-payments and monthly instalments of about 20 per cent of their income. Kundu finds this to be highly questionable in the Indian case. In some parts of the developing world, a section of the urban poor may well have some untapped capacity to save and invest in shelter, given the right assistance and circumstances. Yet, it may be wrong to put pressure on poor people to reduce their food expenditures in order to pay more for housing. The same argument can be used regarding some non-essential consumption items which may give the poor great pleasure.

There are several credit schemes and subsidy arrangements for housing. The largest housing-finance agency is HUDCO (Housing Finance and Urban Development Corporation). It was established in the mid-1970s as a government agency under the Companies Act - to provide financial support to state governments, housing boards and local authorities for implementing housing and urban development projects. HUDCO has been the leading agency of the Government through which the major portion of public investments in housing has been channelled. The agency has a specific responsibility for improving the housing conditions of the lower-income group and the houseless. It works through different public institutions and does not lend to individuals. Expenditures on urban infrastructure constitute a small part of its activities. HUDCO receives funds from the Government as well as from foreign donors. The period of repayment and rate of interest on its loans vary from project to project depending on the income level of the beneficiaries. An emphasis is placed on cost recovery. No subsidy is thus involved, making this agency of less direct value to the poor. The institutions, through which HUDCO works, may, however, assist the poor by cross-subsidizing land and shelter projects. The development cost of an area is then mainly borne by those who buy the high-cost houses. HUDCO has fixed ceilings on the cost of housing for the poor and lower-income group, securing that the quality of materials etc., discourage a hijacking of the units by middle-income people. The thrust of HUDCO is now on self-financing schemes in which funds of public agencies are not tied up for long, and on cooperative housing schemes for which individual savings can be mobilized. This is partly based on the experience that the hire-purchase system blocked resources for a long time and the recovery rate under it was poor. Moreover, the Government has realized the inefficient role played by HUDCO in providing finance to low-income groups.

According to the National Housing Policy innovative savings and lending instruments will be introduced to integrate the housing-finance system into the capital market. This will be done by giving the housing-finance institutions access to the funds on a competitive basis, and by permitting the National Housing Bank (NHB) and HUDCO to set up mutual funds for housing. Steps will also be taken to introduce a secondary mortgage system in order to attract funds from a wide range of investors, including insurance and provident funds, and to integrate housing-finance with the overall financial system. The planning norms for housing at the neighbourhood level will be integrated with the lending guidelines of NHB and other financial institutions. Appropriate ceilings on plinth area will be stipulated in order to discourage large premises. It is envisaged that the housing-finance system as a whole will become self-financing in the long run, to meet the needs of different income groups and purposes, with longer repayment periods, graduated payments and simplified procedures, whenever necessary, to ensure affordable monthly instalments and larger coverage across different urban areas. It is moreover recognized that capital and interest subsidies for the urban poor must be targeted carefully, and that improvements be made in the housing-finance procedures and shelter-delivery system in order to reduce the cost of shelter for the poor to affordable levels. The subsidized housing programmes - such as the Indira Awas Yojna for Scheduled Castes, Scheduled Tribes and freed bonded labourers - will be restricted to the absolute poor. The housing-finance system will be devised so that it can be flexible and able to respond to a variety of shelter needs.

The National Housing Policy will give priority to the promotion of shelter for the homeless and the inadequately housed, i.e., such groups as:

· Households below the poverty line;
· Families displaced by development projects;
· People who become victims of natural calamities;
· Scheduled castes and tribes;
· Single women and female-headed households;
· The physically handicapped.

The NHB has a somewhat similar role to HUDCO, i.e., it is an apex agency working through sub-agencies. According to its chairman, the Bank has no ideological predilections on public-sector housing. The Bank encourages the public sector to acquire large plots suitable for building large townships. But it has the firm belief that public agencies should not preoccupy themselves in construction. Construction should be done by the private sector and cooperative societies (Business India, 4 February 1991). The NHB provides short-term loans to housing agencies for development of housing land, and long-term loans to individuals for house construction. It provides equity support and refinancing facilities to public and private housing companies.

There are also several private housing-financing institutions in India. What is common among these organizations is that they require security and/or previous savings as collateral for loans. The poor are thus left out. The commercial banks give individual loans under NHB's stipulations for a maximum period of 15 years, at annual rates of interest between 10.5 and 14.5 per cent. Loans from this and other private sources, however, require land titles that are marketable and free from any encumbrances. The insistence on security for loans has been the major hurdle for the poor and also for many among the lower-income group. Housing agencies have - by the financial and administrative stipulations guiding their activities - excluded the poor. The requirement of, for example, a permanent address and an acceptable guarantor have become major impediments for the poor to take advantage of the formal housing schemes. Central and state government departments undertake housing for their employees, but not many among them are below the poverty line. Moreover, the urban poor are unlikely to organize themselves into housing cooperatives, and thereby benefit from loans and assistance extended to these. Housing agencies must now be financially efficient. Many of them must therefore abandon - or at least drastically reduce - their schemes for the poor and lower-income group, in favour of the more commercially viable schemes.

Since the cost of housing normally becomes higher than originally planned, and because of the time spent by completing the elaborate administrative procedures, and that delays in construction normally occur, the poor get priced out of even the most generous schemes that are designed for them. The urban poor in India are unlikely to spend more than 10 per cent of their income on housing (Kundu, 1993). Hence, owing to the design of the housing schemes even under favourable circumstance, they fail to reach the poor. The subsidies intended for the poor thus flow to people with higher incomes through regular as well as irregular property transfers.

An overview of existing agencies and institutions (Kundu, 1993) reveals that they do not take appropriate consideration of the demands and affordability of the poor and lower-income group. Many of the organizations exclude the poor from the purview of their activities by their administrative and financial stipulations. One example is the Life Insurance Corporation. Funds from this organization benefit the middle- and higher-income groups. The requirements for obtaining loans from finance institutions are so stringent that housing credit is virtually unavailable to the lower-income group. Even with the relatively low interest rates charged by HUDCO on shelter loans to the poor, the monthly instalments are beyond the affordability of a large part of the target group. Government financial resources for shelter are limited in India, and can only reach a small fraction of the urban poor. The only way of reaching some of the poor at all is to allow for collective guarantees of loans. It is thus essential that CBOs are legally recognized. In view of the failure of the Government's financial institutions to reach the poor, it is necessary to encourage the establishment of housing savings societies catering for the poor by providing small loans with low interest that are repayable over a long time period.

The size of grants and loans from the Central Government and state authorities to the public housing agencies have been reduced recently. The public housing agencies like, for instance, HUDCO, have thus been forced to rely more on their own funds. This has contributed to the policy of self-financing projects. Moreover, the willingness to allocate funds to projects for the poor has weakened. This is due to the fact that the recovery rate on loans to the poor and the lower-income group is less than 50 per cent. The decline in shelter funds for these two groups is also caused by HUDCO's stipulation that 55 per cent of all loans must be targeted to these groups. The development authorities of the big cities are thus somewhat reluctant to approach HUDCO for funds.

It is evident that a considerable amount of subsidy has flowed into the housing sector over the past few decades. It is difficult to assess the percentage given to the poor because of lack of data on total subsidy per dwelling type and the illegal transfers of properties. Nevertheless, an analysis of the functioning of these agencies and the financial and administrative requirements on beneficiaries, make it clear that only a small number of urban poor could have benefited. To ensure that the subsidies reach the poor, a variety of means have been adopted, such as issuance of identity cards to the slum and pavement dwellers. Furthermore, subsidized services have been channelled through voluntary organizations. Yet, experience has shown that local power groups tend to intervene by buying or outright taking over provided facilities, shelter and loans.

During the 1960s and 1970s the funding responsibility was mainly on the Central Government. In the 1980s this responsibility was shifted to state governments and local authorities (see box 42). Many programmes are now facing financial problems that have led to slow progress and implementation. This has adversely affected the availability of basic services for the urban population, particularly the poor. Some programmes, such as the UCD, have been terminated by state governments due to lack of funds. In programmes supported financially by the World Bank - such as SUP and SIP - cost recovery has been sought by increasing water and property taxes. Many local authorities have found it difficult to comply with this requirement. Experience with EIUS shows that the ceiling on per capita expenditure is inadequate for the services to be provided. Even in those cases where improved sanitation has been provided with a SO per cent subsidy, the poor cannot afford to participate. The recovery rate of the investments is accordingly very low. It should, however, be noted that when NGOs are involved, the recovery rate is somewhat higher. Furthermore, in some of the poorest communities the maintenance of the basic amenities provided is not paid by the residents. The local authorities have thus been forced to bear this cost also.

Box 42. Dharavi redevelopment

In exceptional cases the Central Government provides funds for the provision of basic services to the poor. To improve the Dharavi slum in Bombay, a Prime Minister's special grant of 1 billion rupees was given. This project is multi-sectoral, consisting of urban renewal and slum upgrading. Land leases were provided to the beneficiaries, as well as housing loans of Rs. 5000 per household, that are repayable over a period of 20 years, with an annual rate of interest of 5 per cent. The implementing agency was the Maharashtra Housing and Area Development Authority, among others.

Slum improvement on private land is questionable since the law does not safeguard against higher rents being imposed after improvements. As a consequence, no slums on private land have been included under EIUS. This has left out about 45 per cent of the urban slum population from the purview of this programme.

The general conclusion is that basic infrastructure and services in the slums are typically in an inferior state, and moreover, very difficult to improve and maintain in a cost-efficient way. Regarding house building and shelter for the poor, the responsibility has been passed on, to a large extent, to the people themselves, reducing the burden on the public exchequer. An example of a scheme combining grants and popular participation is outlined in box 43.

Box 43. Land and Infrastructure Servicing Programme

The Land and Infrastructure Servicing Programme (LISP) is part of the large World Bank-supported Urban Development Programme in Bombay. The project focuses on developing cheap site-and-service land for housing of different price classes in several locations, in and at the outskirts of the city. Each area should be relatively small. Full cost recovery is emphasised to achieve replicability. The project intends to include 88,000 plots altogether. 60 per cent of these will be made available to the poor and the lower-income group. LISP provides the plots with water, sewerage, electricity and access roads. The plots for the poor are grouped in clusters and the plinth for the toilet is granted. A loan is given for the construction of the superstructure of the dwelling. LISP organizes building permits for a typical house, thus relieving the poor from individually applying for such permits.

The allotment for all plots is made by a computerized lottery. The beneficiaries are then organized in CBOs according to clusters. These CBOs are responsible for the maintenance of infrastructure and for assisting the individual members in shelter construction. A social development group in LISP gives professional advice to the CBOs. The monthly payment for loans and plot service charges should be only 15-18 per cent of each beneficiary's income. The down-payment does not exceed Rs 1000 and the interest rate is 12 per cent per annum. Only 1 per cent of the project cost for technical assistance and training is not to be recovered.

A most important aspect of this project is that people are not moved by the authorities. Individuals apply for a plot after they have assessed the suitability of an area. Then they may be lucky at the lottery draw. This procedure results in a reduced incidence of reselling.

It is probable that slums will continue to grow and squatter settlements will continue to spread in Indian cities, at least in the near future. The significant progress in the agricultural and industrial sectors since 1950, particularly the substantial diversification of the industrial structure, should allow increased investments in housing. At present, housing contributes as little as 3.5 per cent to the GDP (Pugh, 1990). Its potential contribution to economic growth is much greater. The unavailability of land, building materials and housing finance explain this meagre contribution.

4. Nigeria

The existing housing-finance system in Nigeria is severely underdeveloped, and thus ill-equipped to mobilize savings and channel investments to housing. Although a Federal Mortgage Bank has been created to serve as an apex institution, the other components of a mortgage system, such as primary mortgage agencies (building societies, housing associations, credit unions, housing cooperatives etc.) have not become widespread. The critical absence of mortgage capital has had an untold negative impact on housing production. Furthermore, in the absence of life insurance policies or other securities, no institution will finance land purchase. Thus, only very few can make or buy a house with funds provided by formal institutions. The overwhelming majority is financing their housing by personal savings and through assistance from relatives. The poor have thus few prospects of obtaining a decent dwelling.