Acknowledgement

1. Introduction

2. Building the nationstate and centralization of government

2.1 Impact on local urban government

2.2 Economic and physical development policies and their impact on the poor

3. Disillusion and the rise of socialist approaches to development

3.1 National government intervention in local government

3.2 Public sector intervention in low-income housing

3.2.1 The public works tradition of government-built housing and slum clearance

3.2.2 The organized self-help movement

3.2.3 Sites and services and slum and squatter settlement upgrading

3.3 Emergence of NGOs

3.4 Entrenchment of the informal sector

4. Free markets and the retreat of government

4.1 Influence of international agencies and donors

4.2 Mainstreaming the "enabling" paradigm

4.2.1 The managerial/political case

4.2.2 The social/developmental case

4.2.3 The economic case

4.2.4 Vehicles of support

Chronology of influential events

International

India

Indonesia

Sri Lanka

Viet Nam

Notes

Comments on the paper

3. Disillusion and the rise of socialist approaches to development

The capital-intensive forms of development of the early 1950s and 1960s ensured that only those who controlled the modes of production or had access to capital benefited. This left a sizeable portion of society in poverty. A powerful clique developed comprising politicians, bureaucrats and the rich which made decisions for the rest of the population. In countries with strong military traditions, army officers either formed or joined the cliques.

At the same time, the cold war was at its peak with both capitalist and communist blocs fighting for the hearts and minds of the people of Asia. Thus, as inequities in the distribution of wealth became more and more evident, and as awareness of socialist approaches expanded among the educated middle and lower middle classes, disillusion with existing development models and with the role of government started to develop. In the socialist countries, the excesses of collectivization of agriculture and industry, as well as forced attempts to rapidly modernize the economy through capital intensive heavy industry, also led to disillusion among the people.

Consequently, an increasing number of intellectuals started questioning the role of governments as agents of progressive change. This led to demands for redistribution of wealth and a greater role for people in determining policy. However, the cold war caused many unpopular regimes to be propped up, mainly to contain communism. As a result, anti-government armed movements were born, particularly in South-east Asia. The most successful of these led to the reunification of Viet Nam and the demise of unpopular regimes in the Lao People's Democratic Republic and Cambodia. In other countries, a cascade of lesser events emphasized the breadth and depth of popular disillusion. These included the resignation of Ayub Khan in Pakistan and the secession of Bangladesh, as well as the electoral defeat of Indira Gandhi in India and her subsequent trial for abuse of power. In Thailand, the student riots of 1973 and 1976 proved pivotal as were the democracy wall movement in China and the trial of the Gang of Four who had instigated the worst excesses of the Cultural Revolution. The race riots in Malaysia and the communist insurgency in the Philippines also epitomized this sense of disillusion. This era led to the rise of pseudo-socialist governments in capitalist countries which tried to gain greater control of their economies through nationalization and populist rhetoric and policies. In communist countries, slightly greater decision-making powers were provided to lower levels of government, particularly in economic matters.

3.1 National government intervention in local government

In response to these broad societal changes, national governments began to try to develop cities on more equitable lines. However, because "equitable" development models were inspired by centrally planned socialist countries, national government interventions at the local level further exacerbated already weak conditions of local governments. The chief means of these interventions was the semi-autonomous parastatal agency or quasi non-governmental organization (quango) that effectively bypassed local government. A range of such agencies responsible for different aspects of urban development were established throughout the region.

These agencies took several forms, including special project offices that were set up to execute a particular project and then disbanded upon its completion. Invariably, such project offices were for the design and execution of internationally funded projects and were staffed by international consultants who left both the project and the country on its completion. The expatriates usually had local counterparts seconded from other government agencies or departments, as opposed to the local government department that should have been responsible for the project. So the counterparts left at the same time as the foreigners, thereby undermining any possibility of continuity. These circumstances were not confined to projects run by foreign consultants. For instance, the design and supervision of the World Bank-funded North East Lahore Sites and Services project in Pakistan was undertaken by a special project office contracted out to a local firm of consultants. However, this firm was perhaps even more distant from the municipal corporation and the Lahore Development Authority than any overseas consultant could afford to be. As a result, not only did the project drag on for many years, but it incurred substantial losses.

Another frequent problem arising from the exclusion of local authorities from the design and implementation of urban development projects was the assumption that local agencies automatically had the capacity to absorb the routine management and maintenance functions once the project was handed over to them. Often this was not the case. For instance, the Tondo Foreshore Dagat Dagatan Development Project office was set up in the Philippines National Housing Authority (NHA) specifically to manage the upgrading of the Tondo Foreshore, with the intention that, once complete, it would be closed. However, the project office had to be kept on as a separate management office in the NHA long after the formal completion of works because the metropolitan authorities were unable to absorb the loan recovery and infrastructure management functions that had been planned for them.

The most significant of the quangos were the urban development corporations (UDCs) or authorities (UDAs) that were established in many cities in the region in the 1970s. For many, the model was the Singapore Housing Development Board which in 1960 emerged from a transformation of the Singapore Improvement Trust which had been in existence since 1924. This style of development authority had a significant capital budget and powers to borrow on the open market. It could also acquire and develop land and enter into partnerships with, or compete with, the private sector. Many UDCs were intended to provide project design, finance and coordination services only. Other agencies, including local government, were to implement development projects.

Another rationale behind the creation of UDCs was that urban growth had exceeded municipal boundaries so that supra-local bodies were needed to develop and plan cities from a regional perspective. However, the more powerful and successful UDCs, frustrated by the inefficiency of the implementing agencies and the complexities of coordinating them, soon took over the implementation and management of projects and programmes as well. A well-known example is the Calcutta Metropolitan Development Authority (CMDA) which was established in 1970 with a staff of 40 and a mandate to coordinate, finance and supervise the implementation of projects by a total of 89 different municipal and state agencies. Because of its inability to co-ordinate these agencies, CMDA gradually took over project implementation and management functions so that by 1985 its staff had grown a hundredfold to 4,200, divided between nine operational directorates.

At the other end of the scale, however, some UDCs were little more than town planning departments with their responsibility confined to the determination of land uses and the exercise of development controls. Nevertheless, whatever their mandate, UDCs were constitutionally independent of elected local government and in many instances operated with little, if any, reference to it, despite performing many of its traditional functions.

There can be no question that the better resourced UDCs increased the overall level of formal urban development and enhanced the role of appointed public sector professionals. However, in doing so they also contributed substantially to undermining the already weak status, power and skills of elected local government and seriously inhibited processes of participation and local control of development.

Another aspect of this process was that, in addition to the establishment of quangos, central governments in several countries in the region transferred the daily management of local services from municipal authorities to central agencies, rather than attempting to strengthen local capacity. For instance, in Malaysia, federal and state government agencies have assumed direct responsibility for the delivery of local services. In 1975, the Metropolitan Manila Commission was appointed by the nation's President to coordinate the management of local services including fire-fighting, garbage collection and traffic control but it was not long before it assumed de facto responsibility for the daily management of these services.

The strengthening of local government has also often been thwarted by lack of coordination between central government ministries as well as the division, rather than integration, of responsibilities that managerially cohere at the local level. For instance, in the 1980s, the Government of Nepal created a special Ministry of Housing and Physical Planning that was separate from the Ministry of Local Development. An uncommon exception to the usual separation of local government from the major line ministries such as works and housing occurred in Sri Lanka in 1977 when a very powerful Ministry of Local Government, Housing and Construction was created.

Thus, by the end of the 1980s, when pressure was growing throughout the region for both economic and administrative reform (structural adjustment) based on deregulation, decentralization and the devolution of authority to locally accountable bodies, many city administrations were in a weaker position vis-à-vis the rest of society than at any time in the preceding 30 years.

3.2 Public sector intervention in low-income housing

Direct public sector intervention in urban housing grew directly out of widespread and growing disillusion with the initial patterns of national development in the region. It thus began in most countries in the 1960s and 1970s and represented perhaps the single most important attempt by governments to make cities more equitable. It comprised public provision of housing finance, the development of land, or the construction of dwellings for rent or sale. Ministries of housing and government housing departments and agencies were established for the purpose. For instance, in India the majority of the state housing boards were set up in the early 1960s while the Housing and Urban Development Corporation (HUDCO) was established as a "second tier" national housing bank in 1972. The Indonesian National Housing Corporation (Perumnas) and the National Housing Policy Board and Mortgage Bank were constituted in 1974. In the same year, the Thailand National Housing Authority was established as a consolidated public housing agency by merging the Welfare Housing Office of the Public Welfare Department and the Slum Improvement (clearance) Office of BMA.

However, for the reasons just described, public low-income housing was generally provided by parastatal agencies through programmes that failed. Consequently, the history of direct public sector involvement in urban housing in the region has been that of the initiation of ambitious housing programmes for low income groups, followed by the gradual withdrawal of government agencies. This process of apparent retreat has been occasioned not only by governments' inability to meet their construction targets, but has also been in response to changes in the understanding of the role of housing in urban social and economic development. It can be characterized more positively as a three-stage sequence of increasing involvement of individual households and communities in the production of officially recognized housing, leading eventually to the current "enablement" paradigm of support-based partnerships between government, communities and individual households. At the risk of gross over-simplification, this sequence can be identified within the last three decades as follows:

• the public works tradition of government-built housing and slum clearance programmes that, in the sequential model presented here, is most readily identified in Asia with the post-independence period of the 1960s;
• the organized (or aided) self-help movement that was strongly promoted in the late 1960s and early 1970s;
• sites and services projects and slum upgrading programmes that got under way in the 1970s and continued throughout the 1980s in most parts of the region.

A brief review of these stages of policy development is in order before starting a discussion of the "enablement" paradigm because it provides a useful introduction to the current state of the art. It should be remembered, however, that all stages are still current policy in different countries of the region. Thus, although the passages that will follow shortly are written in the past tense as though the strategies they describe form a sequence that is now behind us, they are in fact in one form or another still very much part of the present situation.

Before beginning the review, it is important to realize that prior to these policies, governments throughout the region regarded the production of housing for ordinary people not in government employment as a private sector affair. Such activities were influenced only indirectly by government through programmes that regulated in some form or other the investment of resources in the development of land for residential use. The most common forms of such intervention were:

• land use zoning and development controls to secure orderly and compatible land use and to control the environmental quality of different areas of cities;
• property taxation policies to generate local public revenue on the basis of notional differentials in the distribution and consumption of urban services;
• rent control legislation to fix rents charged by private landlords so that low-income households could find affordable housing;
• controls on the extent of individual property holding in urban areas in order both to redistribute the property of large urban landowners to their low-income tenants and to reduce the extent to which landowners could profit from exploiting the demand for housing by lower income groups. This measure was confined to India, Nepal and Sri Lanka.

Inevitably, the impact of such measures on the housing that was built, particularly by the lowest income groups, was determined by the extent to which local authorities were able to enforce them. Thus, while in upper income areas the land use regulations and building controls were relatively easy to enforce, this was not the case in the large and growing low-income neighbourhoods where such public controls were and still are virtually impossible to police. Such was the regulatory framework at the beginning of the 1960s. Bearing in mind the caveats already expressed, subsequent government intervention has broadly followed a three-stage trajectory.

3.2.1 The public works tradition of government-built housing and slum clearance

The first stage, often referred to as "conventional" housing policies, stemmed from the new political need for governments to be seen to intervene in the housing market in support of the lowest income groups. It was also due to a genuine concern for the orderly physical growth of cities and the appearance of the urban building stock. The aesthetic homogeneity of residential areas, to some extent a legacy of the postwar modern movement in architecture, became a symbol of public affluence, good health and social well-being with which governments and city administrations wanted to be identified. It was genuinely believed that governments could provide subsidized housing for all but the very poor.

The result was the establishment of new public housing agencies or the expansion of existing ministries or departments of public works at the national (or state) level. Virtually no public housing authorities were established at a municipal level. Their first task was to set or adapt standards of space and construction that defined a "minimum standard dwelling" that was deemed acceptable by the professional and political staff of the agency. As these officials belonged to the middleclasses, the standards, although reduced, were still more suitable for middle class steady income earners rather than the poor. These standards became statutory norms for the production of new housing against which the existing urban housing stock could be measured in order to establish the extent to which it needed replacing, for example, "a minimum of 25m² of habitable living space, of permanent construction with direct access to a supply of potable water and water-borne sanitation".

The outcome of this exercise, together with estimates of residential overcrowding, constituted a notional "housing deficit" which, when added to projections of future population growth and the formation of new households, provided an arithmetical figure of "housing need". To this calculation was applied an estimate of those present and future households that could not afford even the "minimum standard house" at private sector market prices: typically a very large proportion. This became the basis on which targets were set for the production of subsidized dwellings by the government for the lowest income groups. Such targets were rarely achieved.

Box 3. Housing production targets - India

Such public housing programmes are typified by tenement blocks of minimal-sized apartments or individual single storey dwellings of relatively high standard permanent construction with individual utility connections. They were commonly located on the urban periphery where land was available and cheap, but were therefore far from centres of employment and social amenities and with only tenuous and costly transport links. They were designed by government architects and engineers whose aim was to produce the lowest cost structures that could meet both the standards set by the by-laws and the professionals' view of "how the urban poor should live". There was rarely any attempt to study the particular needs of the intended users, let alone consult them. The beneficiaries, who were officially qualified by having incomes below an established ceiling or who had been displaced by a slum clearance programme, had no part in the decision making that determined the location, design, standard of construction or management of their housing. There was therefore little chance that it could respond to the individual needs, demands or aspirations of any of its occupants, and no chance that it could respond to those of all of them.

Official controls very often extended to the use of the dwellings themselves, for example: No commerce; No tenants; No animals or market gardening; No extensions or modifications to the building. These arbitrary restrictions were placed upon households that were invariably dependent on being able to supplement small and irregular incomes through such activities, not only in order to feed and clothe themselves, but also to pay for the housing whether it was allocated by hire-purchase of an eventual freehold or rented on leasehold. And, despite the subsidies that were built into the housing, many occupants could not afford the rent, even though the housing was supposedly designed for them.

A major consequence of this was that many housing units were sold or transferred by their intended beneficiaries to wealthier households for whom permanent accommodation had a higher priority, either as a home or as a capital or income-earning investment. The official reaction to this perfectly rational behaviour was frequently one of "moral outrage" couched in terms of the "ungrateful and mercenary" response of the urban poor in using public subsidies ("government charity") with which to speculate. Rarely was it understood or accepted that for low-income households living close to the breadline, the responsibility for real estate was often low on a list of priorities for survival, particularly when a subsidized dwelling represented a valuable asset to exchange either for a lump sum or for rent.

Box 4. Missing the mark: public rental housing in Bangkok

In situations where resale, transfer or subletting were uncommon, usually in rental housing estates which had been cheaply built to save capital costs, environmental conditions tended to deteriorate very rapidly. To a large extent, this stemmed from the occupants' exclusion from any direct involvement in the design of their dwellings, and the consequent perception that they had no responsibility for the maintenance and management of their homes and the common spaces around them. This responsibility was seen to rest with the landlord: the housing authority. However, the public housing agencies were unable to fulfil their management and maintenance functions owing to a shortage of resources. Thus, new, high-cost, slums were created very rapidly. For example, many of the tenement blocks built by the Tamil Nadu Slum Clearance Board in the 1970s were already classified as slums by the late 1980s. Together with more "traditional" slums and shanties, they were high on the priority list for slum improvement by the Board itself.

A slightly different aspect of low-income housing provision was slum clearance. Although the two activities often went hand in hand, slum clearance achieved its own rationale when governments saw it as their responsibility to rid cities of the unhealthy and unsightly slums and shanty settlements that were springing up at an ever-increasing rate. Slum clearance programmes usually concentrated on the removal of self-built shanties instead of dealing with overcrowded, run-down central area slums in old buildings which presented much more difficult problems involving complicated ownership networks and issues of design and construction in or close to central business districts. A notable exception was Bombay, which in 1970 set up a Buildings Repairs and Renewal Board to improve sanitation and structural safety in old central city tenements (chawls). These provided single-room accommodation for thousands of low-income families.

In general, slum clearance programmes solved few problems. They effectively depleted a large proportion of the urban housing stock and destabilized and alienated some of the more vulnerable communities involved in urban development. For example, as Seoul prepared for the 1988 Olympic Games, world attention was drawn to its long history of evictions and slum clearance to little real effect involving millions of low-income citizens. Similar accounts involving the destruction of hundreds of thousands of modest but affordable dwellings abound across the region. In Manila, 90,000 people were evicted and their dwellings demolished in a single three-month period in 1964. A 1985 study found that some 272,000 people in Bangkok were under threat of eviction. In 1975-1977 more than 150,000 people lost their dwellings in Delhi as part of a city beautification programme.

Despite ambitious intentions to rehouse slum clearance victims in new public housing, very few were actually rehoused. Even then, they were often moved to new sites on the urban fringes or beyond where land was cheap and they were "out of site". Such locations were far from centres of employment offering work suitable for semi- and unskilled people who then had to spend a large proportion of their low and usually unstable incomes on transport. In addition, such new low-income housing areas, typically populated by young and migrant populations, were often underserved with basic health and educational facilities. Thus, slum clearance tended to be merely slum relocation as households were forced to start the painful and alienating process of once again illegally setting up their homes in a different place, while waiting for the next round of slum clearance to catch up with them. There were occasional reversals of these programmes where communities were sufficiently organized or assisted to be able to resist them.

Box 5. Manila's Tondo Foreshore

At a different level, public housing projects put a major strain on the construction and building materials industries. These were already under pressure from other national and urban development efforts. This strain was aggravated by the perception that investment in subsidized housing for the lowest income groups was not economically productive. At best it was classified as a politically necessary social overhead. Even in mixed economy countries, it was not considered to be a basic welfare function of the state, such as health care or education. There was therefore constant pressure to reduce the costs of public housing programmes or to curtail them in order to release resources to the more obviously productive branches of the construction sector such as civil and agricultural engineering, transport and industry.

There were two common responses to such pressure. These were either to reduce the subsidies or to cut the costs, or both. The reduction of subsidies meant recovering a greater proportion, if not all, of project costs from the beneficiaries. But this, in turn, meant accepting higher income groups as beneficiaries, effectively excluding the previous lower-income target group from the project. For example, in 1988, the Indian Housing and Urban Development Corporation (HUDCO) merged the two lowest income categories in its classification system for loan eligibility, thereby effectively releasing state housing boards and development authorities from having to construct housing for the poorest groups.

Cutting costs meant either reducing space and construction standards below those previously set and politically accepted as "minimum", or reducing the cost of construction.

3.2.2 The organized self-help movement

The attempt to reduce construction costs was the main reason for the introduction of organized self-help programmes (aided self-help). These programmes constitute the second stage in the sequence of housing policy development in the region. Essentially, the goal was to organize the beneficiaries of new low-cost housing projects into work units to build the project. Although project management varied widely, it was generally agreed that dwellings would not be allocated until the end of the project, thereby ensuring that an equal effort was put in by all to all of its parts. People could thus not concentrate their energies only on the house that would eventually be theirs. As an approach to public housing provision, it was less used in this region than in other parts of the world, notably Latin America. However, it was an important component of the Philippines Land for the Landless programme in Mindoro and Palawan, and in the Indonesian Transmigration Programme in Sumatra in the 1970s. It was also the basis of the Building Together project in Bangkok and the Sri Lanka Hundred Thousand Houses Programme from 1977-1982.

The principal argument behind the organized self-help movement was that by using its beneficiaries to build the project, labour costs could be reduced, thereby reducing overall costs. It was also often argued that, in addition, this approach would obviate the need for private sector contractors, as building materials and construction supervision (the "aid" to the self-help) could be supplied direct from the public works agency. Thus, all contractors' overheads and profits would also be saved. This would, in turn, relieve pressure on the formal construction industry, which could then be employed more productively in other sectors that demanded higher skill and technology levels than the low-cost housing sector.

As for the beneficiary households, it was argued that their collective involvement in the construction of their own homes and neighbourhoods would foster a commitment that would reduce the degree of speculative resale to higher income groups. It would also help to develop a community spirit in an otherwise heterogeneous assembly of low-income families. As a bonus, householders, many of whom were recent migrants to urban areas with few urban skills, would learn a productive trade through the construction of their own houses. Though much was learned and developed from the organized self-help experiments of the early 1960s, the expectations for the approach as a solution to urban low-cost housing problems were shortlived. There were several very basic reasons.

Principal among these was the dependence on complex and sophisticated management processes, not only of often quite large construction sites and sequences, but also of the social interests of participating households. To be at all operable, projects had to start with a period of social preparation which often meant little to the participating households whose ambitions were only to gain legal access to an acceptable and affordable piece of real estate. Even having gone through this process, many rivalries and resentments arose over the extent of one family's labour contribution in comparison to that of others, or because of a period of unanticipated slow progress in construction, or a delay in the delivery of building materials. This often reduced morale and cooperation to the extent of killing the project before its completion.

The aim of cost reduction through the employment of beneficiary householders as unpaid labour also proved dubious. In countries with relatively low wage rates, the total labour component of total project cost of even very low quality permanent construction is rarely, if ever, more than 20 per cent. The unskilled labour component is about one third of the total. Hence, only minimal savings were generated by the use of community labour. In addition, the use of skilled manpower (carpenters, masons, bricklayers, plumbers, etc.) to organize, supervise and train totally inexperienced labour was far less productive than if it was part of a conventional construction team. Moreover, the time available to householders for building rarely fitted the regular working hours of the employed technical, training and managerial staff because communal house construction assumed a lower priority than wage employment, or the search for it.

Furthermore, the obligatory nature of participation in organized self-help was often resented, especially when compared with the fortunes of those who had received fully constructed, subsidized housing at no or little additional cost. For despite their role in digging trenches and laying bricks, organized self-help project beneficiaries rarely had more access to the fundamental decisions over location, design or cost of their dwellings than the occupants of prebuilt blocks of flats.

Even so, despite the problems and failures of the approach as a whole, organized self-help made an important contribution to understanding the link between public housing production and community development. For the first time, the engineers and architects in the housing authorities were joined by social workers. Their duties were to mobilize and manage the work brigades but their work often took them well beyond this into the wider realms of community development and local environmental management.

3.2.3 Sites and services and slum and squatter settlement upgrading

The experience gained in organized self-help, together with several other significant events in the 1960s and early 1970s, came together to launch the third stage in this coarse chronology of direct public sector intervention in the provision of urban housing, namely the development of sites and services projects. Broadly, this approach entailed a division of responsibility whereby government provided those components that could not easily be found or assembled by individual low-income families such as land (sites) and basic infrastructure (services). For their part, each household assumed responsibility for building the superstructure of their dwelling. Thus, for the first time, it was possible for beneficiaries of public housing projects to take responsibility for more than the most marginal decisions concerning the production and management of their dwellings and domestic environment.

The main advances of the period derived primarily from fresh studies by such people as John Turner of the mechanisms by which slums and squatter settlements grew. These, in turn, were driven by the high urban growth rates in Asia in the 1960s which had caused such highly visible solutions to the provision of low-income shelter to proliferate in the first place. For example, in the intercensorial period of the 1960s, the population of Bombay, then a city of some 7 million people, increased by some 900 new households each week while Jakarta and Manila were each adding close to a quarter of a million people a year. By the end of the decade, more than half the population of many cities was living in illegal structures on land to which they had no title. Not only had they received no official recognition or assistance in housing themselves, but they had often been harassed by eviction orders and slum clearance programmes in the process.

John Turner's now famous paper at the United Nations conference in Pittsburgh in 1966 laid bare the mechanisms by which these expanding settlements grew and were managed. It is regarded by many as the starting point of a new understanding of low-income urban settlements. It was followed by studies of informal settlements in many different cities of the Third World in the late 1960s and early 1970s which further developed understanding that, together with a new vocabulary, gradually found its way into official policy.

The main advance was a recognition of the ability and resourcefulness of urban low-income households to produce (or procure) and manage their own shelter and domestic infrastructure. It was demonstrated that home construction is often a lengthy sometimes never-ending process that corresponds to the changing demands and fortunes of the owners and users. Although many squatter households did not actually construct their own dwellings using family labour, highly cost-effective solutions were achieved because they kept a very tight control on the acquisition of building materials and the management and supervision of construction by the tradesmen and artisans they employed. Perhaps most crucially, the importance of security of tenure to land (not necessarily freehold title) as a precondition for individual investment in residential development was reinforced.

At the same time, a change in the perception of the role of housing in the process of urban economic development was beginning to take place, together with refinements in the economic and political arguments for public investment in low-income housing. Not only was the relationship between good environmental health and productivity being made more explicit but, with the development of a better understanding of the "informal sector" as a major contributor to the urban economy, the importance of domestic housing as the site of manufacturing and commercial activity became more apparent. Increasingly it was observed that, given security of tenure to land, even minute household savings were invested in its development. Thus, it was hypothesized that the development of recognized housing in place of illegal settlements would ultimately generate new contributions to property taxes which are a basic source of municipal revenue throughout the region. Furthermore, it was thought that a sense of domestic security would enhance political stability as hitherto disfranchised people gained a recognized stake in urban real estate and hence a commensurate status in urban society.

Nevertheless, these observations and arguments were accompanied by a growing awareness of the extent of environmental deprivation in the expanding slums and squatter settlements. This could not only threaten the health and safety of slum inhabitants, but also those of the formally recognized and wealthier neighbourhoods of the region's cities. Thus, slum clearance programmes maintained a high priority on the official agenda but, in view of the arguments just outlined, they gradually gave way to slum improvement and environmental upgrading projects rather than outright clearance.

The basic ingredients of slum upgrading were the award of legal rights to the land upon which squatters lived, and the provision of access to safe water and waste disposal. Such programmes often went hand in hand with sites and services projects in order to provide new land for those households who had to be moved to clear space for public amenities and use, and safe access.

However, there has been a very wide interpretation across the region of adequate plot sizes and acceptable levels of service provision in sites and services projects. These range from, for instance, the Dakshinpuri project in Delhi where little more than a plot of ground demarcated by four pegs marking the corners and access to communal water and sanitation points was provided, to the Bekasi and Medan projects in Indonesia in which substantial core houses were built by Perumnas. The vast sites and services projects in the World Bank financed Madras urban development projects at Arambakam and Mogapair provided individual water connections and waterborne sanitation to every plot. However, in order to reach the lower income groups with minimal subsidies, the smallest plot sizes were as little as 33m².

Since 1972, the World Bank has played a major role in developing and promoting sites and services and slum upgrading programmes throughout the region. It has in this way extended the principles of replicability (similar projects should be able to be undertaken without outside technical or financial assistance) and full cost recovery (no hidden subsidies) to the shelter sector. Although these measures have certainly brought acceptable housing within the reach of many low-income households previously excluded from the formal market, there are aspects that have been questioned. One of these concerns financial responsibility for the capital cost of service provision. In many sites and services projects, the infrastructure (roads, drains, main water supply, public open space, etc.) installation cost has been recovered directly from the beneficiaries, whilst in the higher income areas in the rest of the city the cost of new service installation has been spread across the whole urban population through the local taxation system (rates). On occasion, in order to design projects that were affordable to very low-income target groups without any form of subsidy, plot sizes have been so small that any future development has been virtually impossible. This was the case in the Madras urban development project already mentioned. The Delhi Development Authority has designed plot sizes of 20.4m².

However, perhaps the most serious, but common, failure in sites and services projects has concerned the assumptions made of low-income households' ability and willingness to pay for housing. Early World Bank-financed projects assumed that some 25 per cent of household income would and could be devoted to housing. This frequently proved to be far too high. Twelve to 15 per cent or even less is a more likely proportion when up to 80 per cent of an urban household's income may have to be spent on food alone. The Indian Housing and Urban Development Corporation (HUDCO) established and applied a working standard of 7.5 per cent for the lowest income groups. However, often, having established a figure for ability to pay, projects have been designed that consume the whole of this amount in the cost of land and infrastructure, leaving nothing with which to build the dwelling. Even where the use of traditional and secondhand building materials is permitted, the construction of a house still requires substantial capital and recurrent payments. And in many projects the design of the dwellings and construction standards were stringently controlled. The use of impermanent and second-hand materials was not allowed for fear that a new, government-sponsored "slum" would be created.

Box 6. Karachi Development Authority's experience in sites and services

Notwithstanding this array of problems with many of the first generation of sites and services and slum upgrading programmes, the approach represented a significant advance. This was particularly so in terms of householder participation when compared to the centralized approaches of the conventional public works tradition and the organized self-help movement. In the most progressive sites and services projects, households were officially allowed to make and implement significant decisions concerning the design of their dwellings: the extent and rate of their investment in the construction of their houses; the process by which they would be built, extended and modified, and who would do the work -- either themselves or a contractor appointed and supervised by them. However, even with these advances, householders still had little if any choice in the location of their housing, the size of land for which they had to pay or the level of services to which they had access.

The beneficiaries of slum and shanty upgrading programmes also gained the security of officially recognized tenure to the land they occupied and, in the best programmes, financial and technical assistance for improvements to their dwellings. Upgrading projects provided a range of infrastructure and services such as water, sanitation, street lighting, paved access ways, surface water drainage and garbage removal. However, residents' participation was generally limited. In many instances, slum dwellers were not consulted on the type or level of improvements that were being installed and were not involved in its management. Thus, it was not uncommon for public amenities to deteriorate very rapidly or even to be vandalized. This occurred in many of the earlier hutment upgrading projects undertaken by the Tamil Nadu Slum Clearance Board in Madras. Quite clearly, "public participation" still meant the participation of low-income households in government housing projects rather than the other way round.

In nearly all countries, sites and services and upgrading schemes continued to be viewed as distinct projects rather than becoming a basis for policy covering all public housing. Indeed, as described in the previous section, special project offices that were independent of the established housing authority or ministry were often set up to manage sites and services and slum upgrading projects.

The Sri Lanka Million Houses Programme, launched in 1983, was the first national housing strategy explicitly aimed at devolving decision-making to the users of public housing. The National Housing Development Authority changed from being a highly centralized design and construction management department to become a decentralized agency for the provision of credit and technical support to low-income households and communities. It was the first coherent experiment in the implementation of the fourth and current stage in this chronology of public housing policies, namely the "enablement" paradigm. This will be discussed in full in the next section.

3.3 Emergence of NGOs

For the moment it is important to note another major product of the period of disillusion, namely the formation of non-governmental organizations (NGOs) and their advocacy of alternative forms of development that were more in tune with the economic and social conditions of the poor. NGOs assumed so many different sizes and forms that a generally acceptable typology has yet to emerge. However, an almost universal theme within the NGO community was and still is dissatisfaction with the status quo. Since this reflects the received opinions of the establishment, NGOs were essentially anti-government in many countries, with those involved sharing a deep mistrust of government.

Within this context, NGO interventions on behalf of the poor have taken two main directions. The first strand was strongly leftist and activist. It opposed the powerful cliques in society on behalf of the poor who were organized to fight evictions, campaign for better working conditions and for redistribution of the benefits of development. It was ideologically driven and often did not focus on actions beyond resisting the establishment. The work of Youth for Voluntary Action in India, Urban Poor Associates in the Philippines, Society of Community Organizations in Hong Kong, and the Korean Coalition for Housing Rights are examples. However, in recent years many such NGOs have realized that cooperation with the government and organization of the poor to meet their own needs is perhaps a more sustainable solution to the housing problem than outright protest. Some have even moved into the sphere of policy advocacy and, at least in India and the Philippines, have been instrumental in changing government policies towards eviction.

Box 7. Fighting against eviction: the case of Bolante

The second strand derived from the tradition of charity. The poor were assisted with the provision of services and infrastructure. The focus was to improve their quality of life through provision of health and education facilities, income generation and credit, and improvement of low-income settlements. While being less paternalistic and more participatory than government, this approach also created a dependency among the poor. Essentially led by middle class activists, this strand was independent of the government but not averse to building collaborative arrangements with it. It has developed towards increasing the capacities of the poor to develop and manage their own situations and also towards mainstreaming alternative development approaches in government policies.

3.4 Entrenchment of the informal sector

Despite all these activities, the impact of government and NGO interventions on the quality of life of the poor has been minimal. In the main, it has been the poor themselves who have evolved development approaches and coping mechanisms to survive and even prosper in cities. Such approaches epitomize free market ideology and have depended on the emergence of entrepreneurs from within their communities. These people spotted opportunities to meet the needs of the poor and filled them through whatever means were at their disposal. Thus, education was provided through home-based tutorials, health was provided by traditional healers and neighbourhood pharmacists, housing was provided by either petty landlords or land-grabbers' Mafia, credit by moneylenders and other vendors.

Most of these development and coping mechanisms did not fit the political, administrative and legislative frameworks of the nation state. They were therefore ignored or made unlawful by the state machinery. The result was the entrenchment of the informal sector. Although meeting the needs of the people, this sector became exploitative in nature because it lacked legal protection. For their part, the poor, with no great ideological considerations in mind, dealt with anyone who could help them meet their needs. Thus, if the local policeman demanded money to allow them to set up a vending stall, they bribed him. If politicians and bureaucrats could protect them against eviction or could provide subsidized services, they voted for them and paid them.

In the past few years, much research has been done on the informal sector, outlining its linkages with the formal and now international commercial and production economies. In many cities, garments and goods for export are prepared by informal sector subcontractors who operate home-based sweat shops. It is estimated that in some cities as much as 40 to 60 per cent of the labour force is employed in this sector. Yet hardly any government recognized its existence or made planning provisions for it. The sustained failure of governments effectively to manage urban areas has led to cities which have physically and socially deteriorated to near chaos including sprawling slums, traffic congestion, pollution and a backlog in the provision of infrastructure and services. In many cities, a sizeable portion of the population feels alienated and the community and neighbourhood spirit is disappearing. There is increasing crime and corruption. In some cities, government control over its functioning has become marginal. Different types of organized crime syndicates, political machines, legal and illegal private sector firms and individuals actually control the functioning of the city and can bring it to a standstill if it suits their objectives.

Box 8. Ismail: the story of a land-grabber