Chapter 10. Selected initiatives on access to land for the urban poor

10.1 Overview
10.2 Sites-and-services schemes
10.3 Settlement upgrading
10.4 Incremental development
10.5 Land-sharing
10.6 Land readjustment
10.7 Housing cooperatives
10.8 Case-studies

10.8.1 Overview
10.8.2 Sites-and-services schemes
10.8.3 Settlement upgrading
10.8.4 Land readjustment
10.8.5 Housing cooperatives

10.1 Overview

In the past Asian cities had a tremendous ability to provide housing opportunities for its growing populations. Land has been available in fairly central location because of the loose structure of the cities and large public landholdings. However, housing opportunities in many Asian cities have rapidly been decreasing as there has been a commercialization of the urban land and rapid increases in land prices. More and more, land is being purchased by highly organized professional land developers with access to a considerable amount of capital. The urban poor have often been forced to move to urban fringe areas remote from income opportunities and often with poor access to public transport. In other cases, the urban poor have to reside on land of poor quality such as hillsides and swampy areas. Housing opportunities for the urban poor has become an increasingly urgent task for municipalities to tackle.

It should be noted that the great majority of shelter is provided by individuals today through private ownership and subletting. Renting as a means of getting access to affordable shelter is becoming more and more common. Most of the initiatives to provide low-income groups access to land for housing discussed in this chapter are implemented by governments. Some issues are relevant for many of the initiatives. For instance, the comprehensiveness of different programmes for low-income groups may vary substantially. Some projects provide many services, whereas some may only provide the most essential component, land tenure. The importance of security of land tenure for settlement upgrading has been discussed separately in chapter 4. Apart from land tenure, the components which are often included can provide physical, social or economic development (table 10.1).

Table 10.1. Possible project components to enhance the living conditions of low-income groups

The core of the matter is income generation. The reason why the rural population migrates to the city is to earn an income in order to improve their standard of living. For rural-urban migrants as well as the larger low-income group born within the city, a rational decision for shelter is determined by the proximity to income generating activities. Turner (1980) recommends that three kilometres from the major employment centre is an appropriate limit when selecting the location for sites-and-services schemes although the actual distance would vary depending on availability of inexpensive transport and the time needed to cover the distance. The basic problem with many low-income housing programmes is the location which have been too far from income-generating activities.

Very few households, independent of income group, can afford to build or buy a complete house using only their savings. Low-income groups rarely have any collateral for conventional loans and they therefore need better access to credit. As an estimated 40 to 70 per cent of the urban population in developing countries earn their income from the informal sector (Choguill, 1994). It imperative to find methods other than conventional banking to provide loan security.

The technology and building material must be appropriate for the specific environmental, social, cultural and economic conditions. In many countries, building standards have been adopted wholesale from European countries and often need to be altered. Furthermore, the price of building material has increased more than median incomes in most countries making it even more important that locally developed materials and building methods are used. The difference in construction affordability between countries was indicated by the World Bank/Habitat Housing indicator programme which found that the production costs for 'median dwelling unit' per one square meter in 1990 was US$97, US$176 and US$145 for countries in the three lowest income quintiles whereas their respective median per capita was US$305, US$655 and US$1505 (World Bank, 1993). These figures indicate that the cost of construction relative to income is much higher in poorer countries.

Subsidies for the provision of housing and infrastructure should only be provided as a last resort. If governments intend to provide some type of subsidy, then it should be well-targeted and provided as a one-time capital grant rather than on an extended basis such as a housing allowance or interest subsidy on housing loans in order to avoid market distortions (World Bank, 1993).

In the discussion of self-help housing, it is also important to bear in mind that some of the basic inequities in developing countries would need to be reformed. To generalize large-scale implications from self-help housing, the "package" would need to include structural reforms on land taxation, zoning and market as well as urban planning policies (Gilbert, 1993). For instance, without lengthy discussions with landowners and detailed planning, many squatter areas could be zoned as high-density residential land. Some experts have been critical of the concept of self-help housing programmes as such:

"By satisfying some of the minimum housing desires of the urban poor, self-help projects can be said to maintain the status quo and as such have been eagerly adopted by urban power élites as multilateral funds have become available over the last years." (Drakakis-Smith, 1976)

The approaches to provide shelter to low-income groups which will be discussed in this chapter are commonly used in most Asian countries. These include sites-and-services schemes and especially settlement upgrading. It should be observed that most of the methods can be described as different sides of the same coin. There are not so many elements that can differ between the approaches and many projects include a mix of approaches. The fundamental difference can typically be found in the approach to land ownership. Apart from sites-and-services schemes and settlement upgrading, incremental development is an approach which lets the target group decide when to develop their land. Figure 10.1 displays the basic difference between these three concepts and conventional housing programmes.

Figure 10.1. Approaches to shelter development

Source: UNCHS (1991)

Governments can carry out almost any task in the shelter development process although they are typically involved in the provision of land tenure and infrastructure. The beneficiaries are likely to have more responsibilities in settlement upgrading schemes and, especially, incremental development. The chapter will also discuss land-sharing which is a method that relies on the principle of developing the land for both commercial and low-income residential use; whereas land readjustment changes the plot borders within an area, some of the profit being used for low-income housing. Cooperative housing, especially condominiums, were increasingly popular as it typically combines considerable influence for dwellers on their housing unit with more intensive utilization of expensive land.

10.2 Sites-and-services schemes

Early attempts by governments in developing countries to provide low-income housing focused on the provision of fully serviced public housing units. Urban migrants and squatter settlements were treated with open hostility. They were generally considered as "dead weight" slowing down the process of development and their illegal settlements were often flattened with the help of bulldozers (Doebele, 1983b).

As it became increasingly obvious during the 1960s and 1970s that attempts to curb rural-urban migration would not succeed and government housing programmes were completely incapable of keeping pace with the enormous demand, there was a growing awareness that alternative methods would have to be found. Many experts (John Turner being the foremost) advocated that if low-income groups were provided security of tenure and, depending on the financial resources of the government, some basic infrastructure, residents would with time gradually improve their housing. It was argued that the role of the government in housing should be changed to be an enabler rather than provider (Skinner and others, 1987). The enabling role has ironically been supported by both 'leftist' and 'rightist' groups, the first mainly because it empowers local community groups and the latter because it reduces the burden on the public purse (Hamdi and Goethert, 1989). The economic advantages with people-based projects were radically summarized by Turner as:

"The bureaucratic, heteronomous system produces things of a high standard, at high cost, and of dubious value, while the autonomous system produces things of extremely varied standard, but at low cost, and of high use value. In the longer run, the productivity of centrally administered systems diminishes as it consumes capital resources, while the productivity of locally self-governing systems increases as it generates capital through the investment of income." (Turner, 1976)

Sites-and-services schemes provide the target group with a plot and basic infrastructure, such as water, roads and sanitation facilities. The beneficiaries either lease or buy the allocated land. Often, they are provided access to a loan with reasonable terms as well as an additional loan for the construction of a house. Although typically not included in the project, it is expected that the plot owner would eventually build a house of reasonable standard. During the 1970s and 1980s, sites-and-services schemes were implemented in nearly 100 countries (Choguill, 1994). The World Bank, for example, implemented 36 sites-and-services schemes (or settlement upgrading projects) between 1972 and 1981 worth more than one billion US dollars and benefitting nearly two million people (World Bank, 1983).

It is difficult to give a description of a "typical" sites-and-services scheme as the interpretation of the concept varies substantially. Some projects only provide pegged-out lots, unpaved roads and footpaths as well as communal pit latrines and water-taps while others may even include paved roads, a utility wall and a partly finished house (Turner, 1980).

There are many physical components to be considered within schemes. First, a utility wall may be provided which includes main service connections such as water supply, electricity and sewerage. Other projects may have a sanitary core while another solution is communal utilities with the option to provide individual connections. Second, the layout of the area depends largely on the planned lot sizes, accessibility to roads and footpaths, and topography. Often the importance of the topography has been neglected. Projects have been built on hillsides and in swampy areas with serious implications. Communal or individual pit latrines also have an impact on the lot sizes and therefore indirectly on the population density. Using individual pit latrines requires physical space for two latrines as one will be in use. Third, some schemes have included the construction of posts and a roof, features which are both expensive and difficult to build. In other instances some walls or a room have been provided as a temporary solution while the household built their house (UNCHS, 1991).

The World Bank introduced two terms as general principles of sites-and-services schemes (Choguill, 1993). First, the term 'accessibility' was used to indicate that the target group (the medium-income bracket of the low-income group) should gain access to the sites in World Bank schemes. However, because of excessive standards and high costs, the targeted group often decided to sublet or sell the site and move back to an area close to his/her original squatter settlement. Second, because of the magnitude of the problem and the limited resources of government, the term 'replicability' was an integral part of all projects. However, for several reasons very few of these projects managed to break even financially (Choguill, 1994). Three out of four programmes failed to recover the incurred costs and, subsequently, the World Bank's support to sites-and-services schemes and squatter settlements has decreased (Gilbert, 1993). None the less, it would be possible for governments to obtain cost recovery from increased property tax revenues.

Cost-recovery has been difficult partly because of the high expenses which allottees typically have to bear early after moving into the area. At this stage, they would have to pay for the infrastructure and construction of the house, while, at the same time, they may either have high transportation costs to their old workplace or have not been able to find a source of income in the new area (UNCHS, 1991). Furthermore, the method has also been expensive for governments as they were typically required to provide land which became expensive, mainly because of land price increases and, to some extent, land speculation.

Many other problems have been encountered with sites-and-services schemes over the years and gradually settlements upgrading has become a more favoured government response to the needs of low-income groups. Sites-and-services schemes have, often, proved to be unaffordable or inaccessible for low-income groups. Some of the problems which have been discussed are:

• Location. Many projects have been situated in the urban fringe in order to reduce costs for purchasing land. These locations have often been too far from income-generating activities. Furthermore, the costs for service provision have increased because of the distance to the existing service network;
• Failure to include one or more components. Many projects have failed to realize how comprehensive a scheme would have to be. Some examples of common shortcomings are access to finance, lack of income-generating activities, failure to mix various income groups (a nearby medium- or high-income area would, for example, provide income opportunities), a healthy mix of commercial and residential land use, social activities such as cinemas etc;
• Selection of eligible households. The process often takes a very long time and is hampered by problems of defining criteria for eligibility in order to avoid corruption and accusations of beneficiaries being selected for political reasons;
• Standards. Government officials in many countries still insist on standards which are too expensive for the intended beneficiaries. Most officials, politicians and professionals have little contact with the actual economic and physical realities of low-income groups and have problems accepting the lower building standards required;
• Administrative delay and shortage of skilled staff. The provision of some necessary services has sometimes been delayed for years because various government agencies have different priorities and lack of coordination. This delay typically increases the costs for the beneficiaries;
• Removal of community leaders. Sites-and-services schemes are likely to remove affluent and influential residents which would leave the squatter settlements without leaders who could press governments for improved services and more assistance (Gilbert, 1993).

10.3 Settlement upgrading

Although the sites-and-services approach offers many opportunities, it is not a feasible method for providing housing to the majority of urban low-income residents because of the huge shortage in the existing housing stock and high costs. Settlement upgrading is based on investments already made in the existing housing stock and is therefore less costly to implement. Settlement upgrading provides existing settlement dwellers land tenure as well as some or all of the components indicated in table 10.1, the fundamental feature being the provision of basic infrastructure. The term "slum upgrading" has been used more widely in literature for the process. However, as some of the areas which benefit from an upgrading programme have housing structures of relatively high standard, the term is somewhat misleading.

In Calcutta, major upgrading efforts during the 1980s included provision of roads, standpipes and sanitary latrines as well as electrification. The Calcutta Metropolitan Development Authority have been instrumental in the programme which is estimated to have benefitted nearly two million people (United Nations, 1986). The Kampung Improvement Programme (KIP) in Indonesia probably rates as the foremost settlement upgrading achievement in the world. The objectives of the programme were to provide access roads, footpaths, drainage, sewage solutions and drinking water and social facilities such as schools and health centres for urban low- and medium income groups in Indonesia's popular kampung settlements. KIP/Jakarta, alone, has improved more than 500 kampungs and provided basic services to about 3.8 million people (United Nations, 1989). Indonesia's five-year-plan for 1989-1994 was to be implemented in 500 cities and included projects for urban renewal encompassing settlement upgrading programmes (Silas, 1992).

KIP is also a good example of the importance of local support for upgrading programmes. When the programme was introduced in 1969 it had a top-down design where officials analysed communities and imposed their solutions. However, as local resistance increased it was realized that the programme had to be reorganized to involve residents in planning, construction and maintenance of the facilities. Funding was instead channelled through community-based organizations. Residents became much more enthusiastic and it was found that they were also willing to provide substantial amounts of funds (Choguill, 1994). Today, KIP is an established method and its effectiveness in the provision of basic infrastructure well recognized.

The process of decision-making is described in figure 10.2. Initially, if tenure security will be provided, it is essential to take precautions at an early stage against the inevitable influx of residents. Low-level aerial photos are relatively inexpensive and would speed up the process considerably as it is then possible to identify the existing dwellings and lots and prevent some of the conflicts.

Figure 10.2. Sequence for decision-making for settlement upgrading

Source: Buranasiri and Tanphiphat (1984)

Regarding the physical shape of the settlement, it will be necessary to decide on the layout. It is possible to produce a relatively regular physical layout within the area which would make infrastructure provision less costly and more efficient but would cause conflicts within the area and destroy much of the existing housing stock. The alternative is minimum movement of houses which causes fewer conflicts and destruction. The solutions differ from case to case and innovative solutions are often required. In Malaysia and the Philippines, for example, it is sometimes possible to move houses since they are often made of bamboo (Turner, 1980). As slums and other settlements, typically, have a very irregular physical layout and houses are closely packed together, it will be very difficult to reduce the risk of fire and to provide sanitation facilities such as pit latrines. If land is very intensively used it may actually be more cost-effective to build a sewerage system rather than to reduce densities to make room for individual or communal latrines. The absolute minimum would, in most cases, be narrow lanes which provide access to specially designed handcarts for garbage collection and fire-fighting (Turner, 1980). In any case it is advisable to prepare calculations of various layout, sanitation and other options so that dwellers can foresee the economic implications.

An evaluation of settlement upgrading projects from 11 countries in different parts of the world made the following conclusions regarding upgrading projects (Skinner and others, 1987):

• Upgrading projects are relatively cost-effective in a situation of high demand for shelter and services;
• Upgrading projects should be simple and down to earth. Basic programmes of service provision were relatively successful, whereas additional components such as income generation and home improvement credit have been less effective. Simple programmes have extended coverage and ensures faster implementation;
• Components to improve land tenure had to be carefully implemented to enhance the perceived land tenure security;
• Increased household mobility appeared to have been exceptional. The process of more well-off groups moving in (gentrification) was found in the Calcutta case whereas there was no clear indication of this type of movement in the Jakarta, Karachi and Madras cases. Regarding the impact on renting households, the rents have typically been increased and, sometimes, the supply of rental accommodation has increased as well. It appeared that the impact on renters varied between the cases and that the impact should be considered at an early stage. It may perhaps be suitable to provide opportunities for renters in sites-and-services schemes;
• Community participation is essential for the success of the upgrading programme although it may be important that the expectations of the upgrading is controlled.

Participatory approaches are important and ideally some form of partnership between the government, private sector and community-based organizations should be set up. Often, however, the success of projects depends on the commitment and efforts of project staff and residents. The need to adjust approaches to each project and to find innovative solutions is universal (IHS, 1993).

10.4 Incremental development

Incremental development can be described as a sites-and-services scheme without the services. The approach includes mechanisms whereby groups of households are encouraged to organize themselves, accumulate funds and to provide infrastructure gradually. Construction begins when the group has collected a certain percentage of the required funds (Peattie, 1982). The approach has been implemented in the United Republic of Tanzania and Zambia as well as Hyderabad, Pakistan.

As with illegal settlements, no infrastructure is provided except perhaps drinking water. The method has the advantage that costs are kept as low as possible thereby allowing access to land for the low-income group. A study of the Hyderabad case provided the following conclusions (UNCHS, 1991):

• The incremental-development scheme reached low-income households;
• While incomes of a sizable number of households dropped initially because of the move to the outskirts of the city, the development of informal businesses in the settlement resulted in an increase in income levels;
• The authorities forced residents to save funds and infrastructure was gradually provided. Furthermore, although there were no standards on housing, it proved difficult to afford both infrastructure and housing;
• High-income groups were successfully barred by requiring permanent occupancy within the area. However, there was scope to reduce land speculation by residents and illegal subdividers;
• The Hyderabad case showed that incremental development was replicable. A comparative study of a nearby sites-and-services scheme and an illegal subdivision, showed that the incremental-development approach was the most viable option (Aliani, 1988). However, Hyderabad was a special case because the demand for land was not so high in the city and the municipality owned vast land areas.

Incremental-development has an inherent problem. Projects bring people together who do not know each other previously and force them to form a community-based organization. The consequence is that CBOs have a tendency to break up after work is finished.

It is obvious that the cost of land in central locations in cities with considerable development pressure will often be too expensive for the low-income group. However, it is also possible to be flexible concerning the payment for obtained land tenure, by, for example, dividing it into several instalments. If the land is still too expensive, then the household would have to find other options in order to find shelter, such as the rental market.

10.5 Land-sharing

Land-sharing is a method to alleviate the housing shortages for the urban poor in Third World countries which has received considerable attention. From the introduction in Bangkok in 1982 until 1994, seven land-sharing projects had provided or were about to provide housing for roughly 6,800 families (ACHR, 1994). In short, land-sharing means that the landowner and the land occupants (squatters or tenants) reach an agreement whereby the land owner develops the economically most attractive part of the plot and the dwellers build houses on the other part with full or limited land ownership. Land-sharing offers several advantages as Governments are finding it increasingly difficult to find land for sites-and-services and other public housing schemes in locations where there are income-generating activities nearby, and eviction is increasingly becoming an unacceptable method to clear land for development projects. The landowner can obtain the most desired land and the occupants can continue living in the area.

The land occupants base their right to the land on two major reasons. Initially, it is common in many rural and traditional societies that occupants obtain a right to the land from living on it and the claim on the land gets stronger the longer the land is occupied. Furthermore, urban development creates land values as a result of investments by the public sector rather than by the landowner. Occupants can, therefore, be considered as having a certain right in the land values as well.

There are four basic characteristics of land-sharing projects. These include:

• Densification. The occupants will be rehoused on a smaller area as the land will partly be developed by the land owner;
• Reconstruction. Densification typically implies that new buildings will replace older structures. It is often necessary to build row-houses to allow higher densities.
• Participation. The transformation of the plots will require a comprehensive negotiation process whereby the community will discuss the allocation of plots and the construction modalities with the landowner, often with the help of a mediator. It is necessary to include all dwellers in the project and to be able to reach agreements within the community.
• Cross-subsidy. External subsidies should be avoided as much as possible. The commercial development should generate a sufficient surplus to cover a deficit resulting from the community's inability to pay for much of the cost of land, infrastructure and possibly housing (Angel and Chirathamkijkul, 1983)

In contrast to other Third World cities, Bangkok has relatively few squatter settlements. As a result of a strict law regarding illegal occupancy, low-income people prefer to reach an agreement with the landowner for temporarily renting the land. Low-income areas in Bangkok are therefore typically slums rather than squatter settlements (Yap, 1992).

The potential for land-sharing appears rather unfavourable in Bangkok, as the law and many politicians support the landowner and many slum dwellers accept the landowner's right to terminate a lease contract and to evict them with no or a small compensation (Islam and Yap, 1989). Partly supported by non-governmental and community-based organizations, slum communities have increasingly become aware of their rights. Community leaders will often risk harassment, fines and being arrested. Whereas negotiations between landowners and communities are used more and more often to solve problems, it is still common for landowners to resort to arson in order to clear slum communities. The reasons for landowners participating in a land-sharing project appears to be either as an act of charity and merit-making or as the last way out of a conflict situation between the landowner and the community which is receiving embarrassing public attention (Islam and Yap, 1989).

Experience has indicated that the chances of reaching an agreement increases if the community can ask a third party to negotiate. It is also imperative that the community stays united in negotiations with the landowner (Islam and Yap, 1989).

An overview of land-sharing projects shows that about 46 per cent of the land was allocated for the slum community on average (table 10.2). In most cases, the land belonged to the public sector and the slum dwellers stayed in the area after land development was completed. It is interesting to note that the type of development (rental or home ownership) depended on other factors than on the amount of land available for the slum community.

Table 10.2. Characteristics of land-sharing projects

Sources: Yap (1992) and Asian Coalition for Housing Rights (1994, Information on land ownership)

¹ Project being implemented.
² The figure without Rama IV is 1.14 hectares.
³ The figure without Klong Toey is 385 households.
⁴ The figure without Inthamara 10 is 294 households.

Some of the problems which have been encountered with land-sharing projects include:

• Availability of land. Often the land available is too small and/or the population density too high within slum communities. Furthermore, this shortage of land may force the building of walk-up apartments which generally are unpopular among slum dwellers.
• Community cohesion. A land-sharing project requires considerable cooperation efforts among slum dwellers who often have a different background. This is particularly a problem during the allocation of plots.
• Complex and time-consuming. The necessity of community participation and agreement throughout the complex process is very time-consuming. The delay in implementation has typically led to increased costs. Furthermore, there is a problem with enforcement as there are no clear rules and each individual household has so far had the powers to block all major decisions. (Yap, 1992)

It can be observed that land-sharing can only take place where slums live under a serious threat of eviction as the community otherwise feels no need for change (Yap, 1992). Although land-sharing can rarely be used because of the many preconditions which have to be met, it is one of the very few methods through which slum dwellers can gain formal access to land without considerable subsidies.

Next: 10.6 Land readjustment