Chapter 7

Urban Land Tenure in Brazil:
From Centralized State
to Market Processes
of Housing Land Delivery

Flávio de Souza and Roger Zetter

INTRODUCTION

Housing policy has always played a central role in the political agenda of successive governments in Brazil. For this reason, the sector provides a valuable barometer, not only of the changing political economy of the country, but also of the link between national priorities and the wider framework of international development trends.

Within this context, this chapter examines recent shifts in Brazilian housing policy and provision evident in the transformation from welfare to market paradigms of development. It presents evidence illustrating how the state's current market-oriented processes and policies (perhaps by neglect, as much as by explicit policy design) essentially endorse or accede to a market-driven process of land delivery, which results not in decreased inequality, but rather in continued insecurity and marginalization for the urban poor. The argument presented here is that the state has failed to promote policies and mechanisms to tackle the growing inequality of income distribution – insofar as this is represented in the housing market. Indeed, the huge differentials between the poorest and the richest members of society are likely to worsen as the state increasingly assumes a peripheral position. Simultaneously, market-led provision of housing land seems equally incapable of tackling the growing social inequalities of the country. This argument is developed through the case study of land tenure security programmes for the urban poor in Recife, Brazil. Some of the principal attempts to develop new instruments designed to tackle poverty and inequality in the housing market will be critiqued.

First, a short contextual overview highlights the contrasting modes of urban-sector policy-making and intervention in Brazil during the contemporary period. The next two sections then review in more detail how recent Brazilian governments have intervened in urban policies – especially housing. The aim here is to provide an insight into the main characteristics and consequences of public policies from a period of centralized state intervention to an era of market enablement. The second of these sections focuses on the changes in housing policy brought about by the neo-liberal development paradigm in Brazil. These changes are presented in the context of the reopening of democracy, followed by a period that we term the consolidation of market mechanisms with respect to tenure issues.

The fourth part of the chapter turns to a specific and important set of policy initiatives adopted in cities such as Recife, designed to enhance housing options for the urban poor through the establishment of Social Interest Development Zones (Zonas Especiais de Interesse Social, or ZEIS). Given the widespread phenomenon of land invasion as the main dynamic of Brazil's urbanization process, the ZEIS programme was designed to serve the dual purpose of promoting land tenure security through the legalization of illegal settlements and upgrading the living conditions of existing informal settlements. By shifting from direct intervention to community-based initiatives in line with the reduction of state intervention, the goal of ZEIS was to secure tenure for residents through partial governmental control of the housing market in these circumscribed areas of the cities. The analysis questions the effectiveness of these initiatives.

FROM WELFARE TO MARKET: AN OVERVIEW OF BRAZIL

Since its inception during the 19th century, the modern state of Brazil has evolved in shape and function, acting to greater and lesser degrees as a regulator of social and economic life. Through its history, the state has played an important role in directing national economic politics through the mediation of internal and external relations and through the promotion of social programmes.

More explicitly, welfare and social policies such as social housing programmes in Brazil, as in most developing countries, have always played an ambivalent role in the country's development. In the case of Brazil, welfare policies were not instituted as the complementary supporting apparatus of the market economy – the case in many parts of the world and, for example, a particular feature of European experience. The ‘welfare state’, at least in developing countries such as Brazil, has had a rather different history. First, it developed not as an antidote to the free market but as an instrument of the military dictatorship with specific ends in that era. Second, welfare policies were inevitably never going to play a supporting role to strengthen the market economy, simply because the Brazilian economy was never powerful enough in its own right. Third, the delivery of welfare-oriented urban and housing policies requires strong local administrations. But Brazilian municipalities have always been very weak and unable to tackle a social policy agenda for many reasons, including their incapacity to collect revenues that could facilitate more socially equitable urban and housing policies. In most cases, the municipalities still rely on the federal transfer of funds. The potential to develop a local social policy agenda now exists, given the strengthened capabilities which the new institutional reforms of decentralization make possible. Yet, despite nearly two decades of post-military democratization, just as this window of opportunity opens, the policy agenda at both national and municipal levels has shifted sharply towards market enablement.

In the contemporary period, and against this perhaps unique background, state intervention in Brazil has been called into question for various reasons. Thus, a notable era of centralized state intervention under the military regime of 1964–1985 has been strongly criticized. It has been argued that welfare policies developed and implemented during that period did not directly target the poor who needed them the most, but rather aimed to promote mass housing construction as a means of supporting the interests of the prevailing elite. Most housing programmes for the poor led to the commodification of their homes, thus opening the housing arena to downward raiding by middle-income groups.

With the downfall of the military regime in 1979 and the subsequent democratic transition period that lasted until 1986, the paradigm of centralized state intervention collapsed and decentralization started to be strongly advocated. It was in this new democratic climate that the foundations for market mechanisms were securely established. Thus, the state's command of development processes has shifted from one of heavy market control to a more laissez-faire approach in which it assumes a peripheral or semi-peripheral role in a predominantly deregulated market system. More generally, through the lens of globalization, the state has been instrumental in reformulating its structure, roles and responsibilities in order for it to function effectively in the new global economic order.

In order to achieve this new configuration, Brazil, like many other countries, has reformed its institutional structures and policies from the very localized level through to the national level. At each of these levels the common, current demand for public intervention is that it adheres to a number of fundamental attributes, which are, among other things, efficiency, flexibility and good governance – in other words, the backbone of the market enablement agenda.

In the field of housing for the poor – the focus of this chapter – new mechanisms to deliver housing and services have been set up. But, as we shall see, their success is questionable.

HOUSING: FROM CENTRALIZED STATE INTERVENTION
TO MARKET SUPPLY

In Brazil, as in most Latin American countries, ‘welfare state’ policies were implemented during the military regime (1964–1985). Before this period, housing land provision was undertaken predominantly through the market, with limited state intervention. Since markets were not providing housing land at affordable prices, the poor only gained access to urban land through illegal or informal initiatives, such as land invasions, plot subdivisions and so forth – a pattern again typical of much of Latin America. Welfare state policies purportedly sought to change that by controlling the land allocation processes from above.

Statistically, the most successful attempt to provide housing in Brazil was, indeed, during the military regime through the promotion of mass housing production. The National Housing Bank (BNH) was created in 1964 with the initial goal to meet a national housing deficit estimated at 8 million units, while redistributing welfare to the families with the scarcest resources and evening out income inequalities (Portes, 1979; Shidlo, 1990a). In total, the BNH produced some 2.5 million housing units over 20 years (Andrade and Azevedo, 1982; Valladares, 1988; Klak, 1990).

The principal functions of the BNH were: to orient, direct and supervise the Federal Housing Bank (Sistema Finaceiro da Habitação, or SFH) and the Federal Savings Associations (Caixas Economicas Federais); to initiate the establishment of savings and their allocation to the SFH; and to regulate housing credit agencies in the national capital market. In this way the BNH hoped to promote ‘the construction and acquisition of houses, especially for low-income groups’ (Batley, 1983, p59). The BNH organized a housing construction programme in the ‘social’ arena,¹ and at the same time was charged with the task of stimulating employment among unskilled and semi-skilled urban workers. The BNH strove to create opportunities for productive employment while alleviating regional and sectoral economic inequalities (Portes, 1979; Shidlo, 1990a).

However, these policies did not actually target the poor, but instead promoted mass housing construction as a means of advancing the interests of the prevailing elite. This happened in two ways. First, because the BNH was required to act as a private enterprise, it invariably satisfied less risky higher-income groups rather than the poor (Melo, 1989; Klak, 1990; Sachs, 1990; Shidlo, 1990a). The poor were provided with barely sufficient shelter, healthcare and services (Alves, 1992). Second, since the middle classes found the idea of home ownership very attractive, ‘this became a central policy of the post-1964 regime in a sense of widening the support and legitimacy’ of the regime in the eyes of the public (Shidlo, 1990a, p36). As such, the BNH, although conceived initially as an instrument for providing low-income housing, became part of a political process whereby the new government could demonstrate tangible results in relation to housing problems (Bolaffi, 1975; Batley, 1983). Of course, the military regime recognized that many kinds of investment were needed in providing state housing for the urban middle classes who supported them (Shidlo, 1990a). But the creation of a federal housing system was the principle means by which the government institutionalized and co-opted middle-class support.

The BNH's funding came from two equal sources: one was statutory, the other voluntary. The former, created in 1966, was the Guaranteed Employment Fund (FGTS): a compulsory savings scheme with an 8 per cent levy on wages to mobilize funds for the provision of low-income housing provided by the BNH. The second source of finance was through voluntary savings from the sale of housing bonds and from savings deposits in passbook accounts within the savings and loans system (Arretche, 1990).

The BNH created a specific financing agency for each of the two segments of the housing market: the ‘popular market’, and the ‘economic market’. The ‘popular market’ was initially limited to low-income sectors where family income was less than or equal to three minimum monthly salaries. This range was later broadened to five minimum monthly salaries: ‘Up to 1975, two-thirds of the BNH's social interest budget was allocated to families with an income range of one to five minimum salaries’ (Shidlo, 1990a, p42).

However, the BNH changed direction and more housing was produced to satisfy the demand from higher-income groups. The inadequacy of the BNH housing policy in targeting the poor was illustrated by the fact that only 15 per cent of the Brazilian population could afford BNH houses. The profit margins for private construction companies, and the high demand for housing by the middle class – as well as their greater ability to pay – redirected the BNH away from the provision of housing programmes for the poor (Maricato, 1987).

Furthermore, a substantial amount of BNH funds was unofficially invested directly in industrial projects, such as the Itapú Treaty, the Tucuruí hydroelectric station, the Açominas steel mill, new metro systems in Rio de Janeiro and São Paulo, and the Carajás iron mining project (Alves, 1992; Shidlo, 1990a). These latter, largely ‘unofficial’, centralized investments in the domestic capital formation of the country were a remarkable overstepping of the wider remit of BNH, identified above, which was to stimulate development and employment. But these activities came to dominate the role of BNH to the detriment of its mainstream housing responsibilities. Nevertheless, they helped to create the foundations and infrastructure for the development of Brazil: this formed the continuing basis for attracting foreign investment after the period of direct state intervention had ceased and, in turn, stimulated further development, employment and consumption.

At the same time, the military government promoted housing production as a way of activating other development opportunities and markets. Housing was a leading sector that increased demand and consumption of consumer durables, goods and services – for example, the demand for public and private transportation to service the public housing estates located in the urban periphery. Moreover, the peripheral location of these public housing schemes in the subúrbios was also the predominant location for most of the new industries. But these industries demanded a more qualified labour force – not the urban poor. Thus, the combined effect of the state housing policies and the industrial location strategies was to consolidate middle-class interests while excluding the urban poor from access to both housing and the expanding urban economy. Even if the housing had been affordable, the dislocation between housing and employment centres for the poor would have imposed additional and unaffordable transport costs on the poor.

The multiplier impact of these new housing estates and the industrial development policies was considerable. But the social and community aspects of the country's housing need were barely tackled.

The BNH faced a financial crisis during the early 1980s (de Melo, 1989; Shidlo, 1990b), which resulted in a reform of the housing finance system engineered by the World Bank. Although the World Bank had financed the construction of low-income housing units in Pernambuco, Bahia and São Paulo during the late 1970s (Shidlo, 1990a), it had tried, unsuccessfully, to intervene more centrally in the housing policy of the BNH during the 1970s. Once the financial situation of the BNH had deteriorated so severely, the World Bank was able, finally, to assume a more central role (de Melo and de Moura, 1989; World Bank, 1989).

Decentralization was a clear objective of the World Bank's strategies, as was the implementation of a structural adjustment programme after 1982 (de Melo and de Moura, 1989). State and municipal housing societies started to look for funding from international agencies such as the World Bank and the Inter-American Development Bank (IADB).

The growing housing deficit in the low-income market, coupled with the absolute growth of the urban population, compelled the BNH in 1983, once again, to reconsider the need to develop housing provision policies for the poor – its original objective in 1964 (FINEP–GAP, 1983). But within three years the BNH closed during another period of economic crisis. ‘Until 1982, the BNH expanded its activities in housing, sanitation and urban development. From 1982 onwards the Brazilian Housing Bank entered a period of severe cuts in its investment programmes as a result of the economic recession’ (Shidlo, 1990a, p43). Most of the BNH's activities were transferred to the Federal Saving Bank (CEF) in December 1986. The CEF became responsible for increasing ‘the rate of public investment in the so-called “social-interest” housing programmes’ (Shidlo, 1990b, p69).

In retrospect, it is possible to argue that state intervention during this period in Brazil continued to serve the prevailing interests of higher-income groups and the elite. In short, democratic norms were not the strong points of the urban agenda in Brazil during the period of the military regime, which was more renowned for its clientelist and elite-driven characteristics. This situation is evident in two respects in the current context. On the one hand, bureaucratic politicking and corruption were employed by officials to support themselves within the existing power structure. On the other hand, by predominantly benefiting those who supported the authoritarian regimes, the controlled supply of housing, among other public goods, provided a powerful vehicle for co-option. Tackling welfare needs, therefore, was not necessarily the main objective in the provision of housing for the poor during this period. Indeed, only those households lucky enough to afford the housing provided by BNH could gain access to some form of official housing, further reinforcing the process of co-option.

Thus, in housing as in other sectors of state intervention, markedly contradictory outcomes were apparent. The official state discourse that extolled the promotion of welfare for the majority differed greatly from the actual policies being implemented, which signally failed to facilitate redistributive objectives. Access to formal-sector and state housing remained limited to those who had the ‘freedom of choice’ of when and where to move house, conferred by affordability. Furthermore, it can be argued that centralization helped to maintain existing power structures and, more importantly, that those same structures were used to consolidate private capital in Brazil, both in political and economic terms. That a centralized state with welfare-oriented programmes should, in effect, consolidate the private sector is a paradoxical outcome not confined to Brazil alone. Nevertheless, it does beg the question as to what might be different under a more avowedly market-oriented development policy, such as the neo-liberal economic paradigm. We return to this question later.

The incapacity of the housing system to service the poor in Brazil was not only related to the inefficiency of state housing provision but also to the state's failure to regard the informal housing sector as a productive element of the economy. Housing policies throughout this period neglected the productive potential of informal-sector housing supply, which, in turn, could not complement the formal housing market, either in the public or private arenas. To the extent that housing policies addressed the informal sector at all, then the government's housing programmes were primarily designed as a means of replacing existing favelas (squatter settlements), and suppressing the further proliferation of these purportedly unacceptable developments in most major urban centres.

After 1985, with the return of a civilian government and the collapse of the centralized housing system, an era of renewed democratization and decentralization began (Souza, 2000). But this period also saw the institutionalization of much more avowedly market-oriented government policies that, although in different ways, have similarly impacted negatively upon low-income urban dwellers.

Most bodies responsible for low-income urban housing programmes in Brazil have, especially during the current period, prioritized cost-cutting, risk reduction and cost recovery over the empowerment of the poor (by way, for example, of providing incentives to the informal sector). Squatter settlements continue to expand in most large urban centres in Brazil, and yet the progressive withdrawal of the state has induced a growing deficiency of formal housing alternatives for low-income households. Self-help housing outside the private market seems to be the most effective way in which the poor can keep pace with the demand for housing. Most of the time, however, these informal housing networks are illegal, and the poor face the constant threat of eviction associated with the lack of tenure security.

Formal housing providers and financiers, including private developers and financing agencies, wield significant power in housing programmes and receive profitable benefits. However, the high price and prolonged delay of public housing supply, as well as the small plot size and poor quality of building materials (which, ironically, are often more expensive than materials available on the open market), have induced some urban dwellers to enter mutirão (self-help housing programmes). Thus, the mutirante (participant of self-help programmes) obtains a house through participatory programmes in which he or she works, not out of choice but out of sheer necessity.

Both municipal and central governments have initiated programmes in which favelados (squatters) are resettled from vulnerable topographies to new, distant suburban areas in the name of ‘health improvements’. These new settlements, however, are often costly to live in and far from local job opportunities and key services. As such, the resettlement solution merely replaces one set of problems with another, such as the crippling impact that high monthly housing payments have on a household's livelihood strategy, and an often sharp increase in household transportation costs. Some of these new areas are considered by the government to be temporary, but many become permanent. In response to these issues, some of the favelados return to the area of residence from where they had been evicted. These locations are often free to live in and are close to employment opportunities. Although this practice predominated during the military regime, it is still a significant element in the dynamics of Brazil's process of urbanization. The Brazilian authorities have tolerated the increasing practice of illegal land development, arguably for electoral reasons and, pragmatically, because illegal land development is, in the end, a housing solution that requires little investment from the state.

Accompanying the process of market enablement and in response to fiscal crises (Melo, 1990), there has been the inevitable process of decentralizing power from central government to local authorities. Brazil has followed this worldwide imperative (Castro, 1991). But an immediate consequence in Brazil seems to have been the swelling of poverty levels as state welfare recedes, compounded by the increasing unemployment caused by market enablement policies. The argument that decentralization might risk increasing social exclusion in municipalities more preoccupied with funding and resources than with public well-being is powerfully made in Brazil, as elsewhere (Melo, 1996).

Moreover, the transition from a period of authoritarian government to an era of freedom and choice appears to be a highly uncertain time for both governments and civil society while new power structures establish themselves. These outcomes further impact upon housing policies and programmes, especially options for the urban poor. On the issue of decentralization and democratization, tensions emerge between new elected municipalities and regional and central levels. For example, efforts to promote democratic rule in municipalities can be found across Brazil (Recife, Porto Alegre, Fortleza, Belo Horizonte and others);² but in regional and central government the stigma of the military regime, in which social policies practised unequal distribution and lacked a tradition of public participation, remains. Accordingly, while organized civil society is now constitutionally enabled to participate in the design and implementation of urban policies, in practice civic involvement can still be manipulated and co-opted by the ruling elites, as during the past, and the urban poor are marginalized.

More specifically, with regard to housing, the shift from a purportedly welfare approach to a market supply system in Brazil has intensified competition that the urban poor confront in the marketplace. By neglect or an explicit policy agenda, low-income urban housing has become scarcer, while disputes about the available housing stock have become more and more complex and entrenched. Withdrawal of the state has enlarged the arena for imperfect, highly competitive markets. In an attempt to safeguard access to housing for the poor, some municipalities in Brazil have introduced housing policies that radically contrast with earlier interventions, although they do mirror approaches to slum upgrading programmes used elsewhere, but with some particular points of distinction. These more recent trends are now discussed.

TRANSITION IN URBAN HOUSING LAND POLICIES
IN BRAZIL

A consistent theme of the discussion so far has been the contradictory outcomes of the ostensibly social and welfare rationale of Brazil's urban housing and land policies. We have seen how rhetoric and reality diverged and how the purportedly welfare objectives of the military regime, in fact, consolidated its middle-class power base. To this extent, the more recent shift to a market-led paradigm that commenced towards the end of the military government during the mid 1980s merely makes transparent what was the reality of the preceding two decades – that the poor were marginalized from state housing policies designed for their needs. This section explores the transition phase in more detail, showing how state instruments and policies to regulate and manage urban housing land for the poor have been influenced by the shift to the neo-liberal development paradigm.

In the context of the economic recession of the 1980s, and the consolidation of neo-liberalism worldwide, the Brazilian state's withdrawal from housing policy is not so much an explicit policy shift as an enforced outcome. The closure of the BNH in 1985 represented not just the withdrawal of the Brazilian government from housing policy at the federal level; it also left state and municipal housing agencies obsolete.

Following the new Brazilian constitution of 1988, municipalities were empowered to set up policies to solve tenure disputes and to ameliorate some housing conditions in existing settlements. But no longer did they have an active role in construction, apart from individual initiatives in Porto Alegre, São Paulo, Belo Horizonte, Rio de Janeiro, Salvador, Recife, Fortaleza, and a few other cities in Brazil. These experiences have had very little impact upon social housing needs, as a whole. Even given that incremental changes in social housing are now considered to be a more appropriate form of intervention than the mass housing construction projects of the past, these incremental changes are still localized and have not yet altered the social construction of land rights reform.

During the early 1990s, at a central government level, deregulation policies and clearly stated pro-market approaches became dominant features of Brazil's political economy. The federal government announced, and for the first time rapidly implemented, political and economic reform measures. In a two-year period, from 1990 to 1992, a remarkable process of market empowerment took place in Brazil (Valenca, 1999b). Social reforms were linked to the ability of markets to deliver services. But the emphasis on public interests was rarely strong. Thus, in the housing sector, the focus of this chapter, finance reforms and unregulated interest levels favoured mortgage-lender groups (Valenca, 1999a). These changes negatively affected the mortgage holders’ ability to pay for their monthly instalments, since their total debt increased more than the market value of their property. Many people lost their properties through repossession because of their inability to repay debts that increased extremely steeply. Even a decade later, many mortgage holders still face civil action to repossess their houses as a legacy of this drive to market enablement. In addition, the role of the SFH (the Federal Housing Bank) was drastically reduced in a number of ways, which have favoured the private housing markets since social housing was meant to become a profitable business as well.

Until the mid 1990s, the social housing sector was still in a chaotic condition; but after the election of President Fernando Henrique Cardoso and the stabilization of the Brazilian economy, expectations were high that social housing policies could once again be social. In contrast, however, during both periods of his presidency (1995–1998 and 1999–2002), little noticeable or lasting effect on the social housing sector can be discerned. For example, there is little evidence of improving access to credit, building materials supply, land delivery, or more proactive policies to support low-income group initiatives to ameliorate their housing conditions, either in the case of existing housing settlements or in new development sites. Social policies existed; but their focus was public health and education, not the urban housing sector. Where positive experiences existed in the housing sector, these have been spread over the few municipalities that have adopted individual initiatives based on a willingness to tackle social housing needs; they have not been part of a systematic national policy.

A potentially significant reform in the context of this chapter was the 2001 federal law known as Statute of the City – Law 1-251/01 (Estatuto da Cidade – Lei 10257/01), which has been regulated by Medida Provisoria No 2220/01 (MP 2220/01). This mandates all municipalities over 20,000 inhabitants to (among other things) review their master plans and incorporate ZEIS areas by the year 2005. These are programmes to upgrade low-income informal settlements and are discussed below in ‘Alternative land delivery systems in Recife: A critique’. This reform will certainly impact upon land markets for low-income dwellers in Brazil in potentially crucial ways, although it is still early for predictions to be made. But particular attention should be paid to the outcomes of this reform and the impacts in housing land markets. An important qualification is that ZEIS declaration alone is unlikely to make a dramatic difference. Other changes are necessary to activate markets and stimulate loan applications in this housing sector, as explained above.

At present, three main housing programmes for the social housing sector remain: Programa de Arrendamento Residencial, or PAR (Housing Lease Programme), Law 10.188/01 enacted in 2001 promoted by the CEF; Programa Morar Melhor, enacted by the federal government in 2001; and Programa Habitar-Brasil BID (promoted by the IADB and the federal government in 1997).

PAR is a loan programme that developers can apply for, provided that the total cost of housing units is set at a fixed price (approximately 22,000 Brazilian Reals – equivalent to US$7000). Residents apply for a lease contract for 20 years, and provided that all monthly payments are paid without interruption, the housing unit can be transferred to the leaseholder. This programme has had some impact in the marketplace since it has attracted some developers to construct in this sector; but little is known about the impact of the programme on public policies to address the mass of low-income group housing needs at municipal level.

Programa Morar Melhor aims at promoting urban development for the regions of highest poverty concentration (complementary to the poverty alleviation paradigm elsewhere). The main goal is to broaden access to sanitation and to improve the living conditions of the poorest urban inhabitants in Brazil. However, again, there is little evidence of its impacts to date.

Programa Habitar-Brasil BID is a programme for municipalities to apply for loans (with the federal government as guarantor) to legalize land tenure in illegal settlements, while at the same time upgrading the physical housing conditions – rather similarly to ZEIS instruments discussed below. In contrast with PAR, this programme is promoted by a multilateral agency (and not by the federal government through the CEF). While, in theory, the impact of Programa Habitar-Brasil BID could be claimed to be an influential policy reform addressing social needs, in practice most municipalities struggle to access the finance, having to demonstrate the familiar market enablement agenda of efficiency, efficacy and managerial capacity to deal with the growing housing deficit of low-income groups.

With the election of President Lula da Silva, who took power in January 2003, another shift in social policies can be seen. This is towards the promotion of poverty alleviation (for example, Fome Zero), consistent with the new rhetoric of market enablement, but in this case almost totally dissociated from housing policies at the present time. The presence of the Partido dos Trabalhadores (PT) (Workers’ Party) in federal government has certainly provided legitimacy for further democratization at this level. However, it is premature to offer an evaluation of the impacts on the social policy arena.

ALTERNATIVE LAND DELIVERY SYSTEMS IN RECIFE:
A CRITIQUE

Over the last four decades, and against the backcloth of an enduringly precarious economy, misappropriated policies and weak institutional capacity, the impact of government intervention in housing supply for Brazil's urban poor housing presents an essentially pessimistic picture – regardless of the contrasting ideologies of the different governments. Fuelled by a process of rapid urbanization (Brazil's urban population increased from 45 per cent in 1960 to over 80 per cent in 2000), and the outcome of inadequate public policies and failed market supply mechanisms to deliver sufficient housing, the critical feature of Brazil's housing production process for low-income dwellers has been informal urban settlement.

It is almost impossible to calculate the number of invasions in Brazil over recent decades, due to the lack of available data. Characteristically, there are hundreds in each of Brazil's large cities and they exist in smaller cities, too, almost certainly in proportion to the overall scale of urban growth. It is estimated that there are over 600 squatter settlements in Recife alone. Since most cities have quite tight administrative boundaries, the number of invasions is inevitably decreasing within cities – since the stock of un-invaded or undeveloped land is declining – while the process of new invasions of informal settlements has been transferred to the peripheral municipalities where land is still available.

By default, this is the predominant urban housing supply system of Brazil – as in much of the developing world. Accordingly, the implications of this process of informal settlements have always figured centrally in government policy-making. The key factor here, given the enormous scale of the process and the failure of public- and private-sector formal-sector housing delivery systems, is the different degrees of legitimacy that these settlements are afforded in the eyes of the authorities, and the different levels of consolidation that exist within them.

The essential policy instrument in this respect is the ZEIS declaration, which is provided for in many of these settlements.³ It could be argued that the ZEIS process as a form of social policy or ‘semi-welfare’ intervention is aimed at resource redistribution – although it is reactive in nature in the sense that successive governments in Brazil have come to accept the need to afford minimum protection to the mass of the urban poor in an otherwise failed housing supply model. From this perspective, the use of the ZEIS instruments merits special attention in the context of Brazil, and the case of Recife is used to explore the impacts and implications.

In Recife, the ZEIS zones form part of a programme in which the municipality of Recife aims at urbanizing invaded land (through the implementation of slum-upgrading programmes and regularizing land tenure) in an attempt to integrate such settlements within the urban grid (Law 16113/95).

Evidence from other developing countries (for example, Zambia and Botswana) confirms that where some form of land tenure recognition is provided, such as through ZEIS, a level of tenure security is created that enables informal settlers to consolidate their housing. The PREZEIS law (Plano de Regularização das ZEIS – ZEIS Regularization Plan) of 1987 (updated in 1995 and again in 1997) has provided the legal framework for regulating disputed settlements since its ratification.

However, there is contradictory evidence on the outcomes in Recife. Against expectations, perceptions that tenure security has been strengthened as a result of the ZEIS declaration appear to conflict with the actual outcome. The extent to which this recognition affects people's perception of their land tenure security is more fully explored in de Souza (1998). But, briefly, the explanation for this muted outcome is as follows. Once an area is designated a ZEIS, the ownership of the land does not automatically change (as in some other countries where provisional rights are created), nor, more significantly, do the households’ perceptions of land tenure security increase.

The efficacy of the transfer of property rights from landowners to possessors (illegal occupants of land) since 1987 has been less successful than the general public expected since the process of reconfiguring land tenure becomes highly contested. Despite the legal provisions, it is subject to endless disputes in court so that the illegal occupiers of land still face the possibility of eviction (see, for example, de Souza, 1998, 1999a, 1999b). The inelastic supply of urban land has contributed to the increase of land disputes among households in search of a house, and has sent the value of the remaining land stock soaring as a market commodity. Much of the time this leads to violent clashes not only between landowner and invader, but also between invaders themselves (de Souza, 1998).

Nonetheless, the ZEIS programme has still been actively promoted in Recife. This is not so much from the viewpoint of enhancing land tenure security – which is not the case, in practice – but as a mechanism to reduce the risk of eviction from housing (as opposed to land), which provides another form of public recognition of illegal settlers’ rights. These objectives are buttressed by the implementation of upgrading projects to improve physical conditions of the settlements. Thus, the ZEIS mechanism has been more attractive for households, as opposed to those actually claiming a right of occupancy on the land, since improvements take place irrespective of the realization of land legalization. The provision of services and other housing resources has caught the attention of other communities wishing to apply for ZEIS status.

At present, declaration of a ZEIS in Recife is the only available mechanism for the legalization of land tenure at the municipal level. In this respect, it appears that the process of recognizing minimal occupancy rights for illegal settlers functions as the only social intervention in housing supply systems in the absence of other more substantive welfare state programmes in Brazil. Even so, the intervention is severely constrained since, in practice, even though public recognition of squatters’ rights under a ZEIS declaration is evidence that the municipality is willing to make these settlements eligible for legalisation, the final guarantee of those rights has to be determined in court. Declaration of ZEIS areas is thus the initial turning point in this process; but the municipality lacks both the overall power to manage and coordinate the policy instruments and the final authority to make the declaration legally binding.

Although the use of the ZEIS instrument has helped to resolve land tenure disputes and to channel resources into upgrading programmes in Recife, the impacts have served to produce both arbitrary and, possibly unintended, structural consequences. This is because the instruments used by ZEIS to legalize land tenure by individual title deeds (on private land) and lease contracts (on public land) have, in effect, become one of the main pillars of instituting formal market mechanisms in squatter settlements. Until recently, the existence of informal housing markets has enabled the poor to compete and sustain, to a degree, their stake in the housing process. However, by deploying the ZEIS programmes, the state has paradoxically played a very strong role in activating housing land markets since the ZEIS declaration commences a process of transition from informal to formal housing market processes. This has begun to attract the better-off to these areas and, in turn, starts the arbitrary process of forcing out the very poorest residents who lack the means to compete in these newly installed markets. Legalization for some thus aggravates the issues of tenure insecurity for others. A key point to emphasize here is that state intervention has played a very important role in changing people's attitudes towards the desirability of legalizing land titles as the main response to tenure security.

Some detailed examples will help to illustrate the nature of these contradictory processes.

A land invasion took place in Recife in January 1991 and the resulting settlement was named Iraque by invaders.⁴ The legal landowner claimed repossession of his land immediately after the invasion took place to avoid the loss of any compensatory payments under provisions of usucapião especial urbano (adverse tenure rights).⁵ Therefore, facing the prospects of eviction, Iraque was rapidly occupied and the municipality of Recife declared Iraque a ZEIS. This inevitably raised the invaders’ expectations of land legalization. But no permanent services have been provided for households (only household waste has been collected), nor has any legalization taken place. Since tenure disputes are time consuming, residents still live in the area. But after such a protracted dispute and de facto permanent settlement, the resolution of the issue is now largely a matter of defining a price for the land so that either the municipality or the state reimburses the landlord to avoid mass eviction of households. The ultimate outcome will, in all probability, protect the invaders, but at the price of commodifying the land.

The example of Entra Apulso – another ZEIS originating from land invasion during the 1960s – illustrates the converse situation where ZEIS declaration fails to protect land invaders from removal as a result of formal-sector land market development pressures. The land market is very active in the area since there has been a growing demand for the construction of middle-income apartment blocks locally, and development of this kind is already featuring in Entra Apulso. Moreover, a number of invader households have been compensated for the loss of their properties caused by the expansion of one of the largest shopping centres in Latin America (Shopping Centre Recife, or SCR). Thus, householders now admit the inevitability of their removal as a result of these market pressures from formal-sector demands, despite the fact that the area has been declared a ZEIS, services have been provided and the municipality has claimed that legalization of land is in the final stage of consolidation and permanency (de Souza, 1998). At present, though, clear titles have not been finally provided, and the massive presence of SCR continually threatens residents with the prospects of removal. This example illustrates how ZEIS designation fails to provide permanent protection against the operation of formal land markets.

The third example illustrates how, in typical conditions, public intervention does little to protect low-income urban land invaders from the inherent vulnerability of their situation. The threat of, or actual, eviction and the emergence of land markets that are just as active as in the formal sector are continuing features of their precarious foothold in the city. Sitio Grande was privately owned land when a peaceful land occupation took place during the late 1940s. However, it was only after the main land invasion took place in 1988 that the then landowner went to court to repossess his land, winning the legal dispute in 1993. Some houses were demolished, politicians and residents clashed with the police, and some politicians advocating on behalf of the residents were arrested. The dispute became highly politicized after these violent events, and the state felt forced to assure residents of their tenure security. The state housing company COHAB-PE acquired the land with the intention to provide clear titles for residents.

At present, no land titles have been issued, and the state authority is unsure about the kinds of land tenure that the residents will gain. Physical disputes still occur in Sitio Grande ten years after state acquisition of land, and many residents have sold their properties or have been evicted by force, sometimes because of the violent conditions of life within the settlement and disputes between neighbours. Without the legalization of land tenure, there are no means of halting tenure disputes or providing dispute resolution. Internal markets have been operating, without state intervention, in similar ways as the private markets. Thus, despite the framework of public powers to protect low-income urban dwellers, a market-driven land delivery process reinforces conditions of inequality, marginalization and insecurity.

Some of these structural contradictions are considered in more detail below. The fact remains, though, that even at a procedural level, the effectiveness of this land policy in clarifying and delivering land titles for the residents of ZEIS is, as we have seen, extremely limited (de Souza, 2004). Clearing titles is time consuming, expensive and encounters resistance on the part of the judiciary in Brazil.

There has been a recent move in Recife towards a proactive social policy named participatory budgeting (PB) (orçamento participativo), where the municipality enables large groups of people, effectively in a commission format, to decide the priorities, objectives and goals of the future investments at municipal level. The introduction of this new policy framework is very much in line with the governance and decentralization reform agenda of the World Bank under conditions of neo-liberalism.

Although there are other successful experiences in Brazil (see Souza, 2000, and Melo et al, 2000, for examples), there is a suspicion that PB priorities may undermine the more precise PREZEIS form of machinery and the objectives and priorities for the ZEIS areas. It is a matter of great concern that programmes for ZEIS areas are not apparently as active as they were during the 1980s and early 1990s. It appears that the new and largely untried but powerful PB instrument might effectively replace and, consequently, overlap with the PREZEIS forum, taking away fund allocation and priority listing from the PREZEIS programme.

In a recent study in Recife and Porto Alegre (de Souza, 2004), just such an outcome was noted. PB was becoming a very attractive instrument for the ruling Labour party in Recife. Some professionals claimed that the PREZEIS forum was perceived to be in decline in almost inverse proportion to the increasing influence of PB in Recife. Personnel who had worked in the PREZEIS forum had been moved to other less prestigious sectors in the municipality and replaced by pro-PB personnel.

Although neither the PB programme nor the PREZEIS forum are intended to consolidate market mechanisms – in fact, their objectives are the opposite – these instruments have perversely worked towards the fragmentation of social welfare policies on the part of the state. While aimed at solving entrenched inequalities, such as access to urban land, these instruments have, in effect, consolidated individual interests and rights in land delivery processes. The PB machinery, in particular, is having the effect of individualizing social interests (although in a democratic arena) by pitching social needs against private interests. The PB process is predicated on the assumption that by putting all interested parties together to discuss development priorities, a balance of interests and power structures could be achieved which would per se guarantee a so-called democratic arena, while safeguarding access to social policy programmes, such as land and housing rights, for the poor. In effect, though, PB, in particular, and ZEIS designation, to a lesser extent, appear to be inviting powerful and influential class interests to dispute the rights of the destitute and the powerless to social policy interventions.

The political redistribution of power, inherent in these apparently participatory mechanisms of governance, was a very influential factor in determining public acceptance. The evidence suggests that, although the initial claims have positively impacted upon the democratization process in Recife, and in Brazil as well, they have also contributed to the increasing withdrawal of the state in terms of its roles and responsibilities to promote social equity and welfare policies in land delivery programmes, such as the ZEIS case discussed above.

Both the PREZEIS forum and PB neglect the recent history of Brazil where local or national elite shave co-opted power structures to sustain their longer term privileges by appearing to sponsor the interests of the poor in the short term. For example, the ZEIS declaration, on the face of it, can be seen as a pro-poor policy since it has retained some degree of social housing rights for the urban poor, although it is less effective than anticipated. However, it can also be seen as a tool for more powerful interests to safeguard a long-term stake in the land market. By changing land designation in Recife, the ZEIS areas can be converted into valuable future assets in the housing land market. In effect, the ZEIS declaration can be seen as a strategy of commodifying illegal settlements by ensuring that such land will inevitably come onto the market once formal legalization is finally accomplished.

The ZEIS mechanisms could be seen to satisfy the twin objectives of acceptability to the advocates of pro-poor policies in Brazil, while being consonant with the global orthodoxy of promoting cities as marketplaces in which land delivery becomes progressively reconfigured as a formal-sector commodity. From this perspective, and despite the fact that ZEIS mechanisms may be receding in importance compared to PB, it can be expected that the ZEIS mechanisms will still retain their place in the policy framework of Brazilian municipalities. Indeed, ZEIS has been enforced by Statute of the City and MP2220/01. But few studies have been produced to challenge the counterproductive consequences and the impacts of this policy in terms of activating housing land markets in the social housing sector.

CONCLUSIONS

This chapter has illustrated how the state's housing and land policies essentially endorse or accede to a market-driven process of land delivery (perhaps by ignorance as much as by explicit policy instruments), which results not in decreased inequality, but rather in continued insecurity and marginalization of the urban poor. The key conclusion to emphasize is that, regardless of the model of government, urban housing policies in Brazil have consistently failed to provide either sufficient new housing, or to improve the existing inadequate housing and land security conditions of the poorest citizens. Thus, the development of more effective housing for the urban poor remains a dominant issue on the political agenda for Brazilian governments.

Centralized housing policies during the military regime (1964–1985) were efficient in terms of housing unit construction. However, lower-income groups were not the beneficiaries of these programmes, which, given the financial parameters under which they operated, were largely targeted at middle-class citizens. Moreover, as we have seen, the peripheral location of these public housing schemes, combined with the process of peripheral industrial development, further reinforced the exclusion of the urban poor. Indeed, the poor suffered not only as the missed target group, but also as victims of slum clearance policies, too. Yet, it was not feasible to ‘clear’ the favelas in Brazil's cities. The state could no more mobilize sufficient resources to build enough new houses than it could resist the relentless pressure of rural-to-urban migration, which merely accentuated the demand for urban housing and, thus, new illegal land invasions and disputes. This was the only conceivable response of the poor in the face of a serious deficiency in adequate and affordable public-sector shelter provision.

In this respect, the chapter has highlighted the structural contradictions of avowedly social policy interventions by the state. Thus, the development paradigm that claims social interest (as is the case with the housing land policy approach exemplified by ZEIS policy), in reality, also serves the purposes of activating housing land markets and putting at risk the survival of the very groups whom it is designed to support: low-income urban dwellers. Clarifying land titles can result in the liberalization of market forces, but this rationale should not be deployed unconditionally – as is now the case – if the social interest is clearly to target the poor who need a secure form of tenure, but which is, at the same time, affordable.

After 1985, decentralization of power from the central government to local authorities and more participatory forms of governance became the dominant approach adopted by the Brazilian government. This transformation, a synonym for the withdrawal of the state from social policy fields such as housing, was consistent with the global agenda of restructuring the role of governments and decentralizing state interests stimulated by the fiscal crises, increasing unemployment and poverty (Castro, 1991).

Conjointly, the poor seem to have found participation to be the most effective way of acquiring formal access to housing in Brazil. Thus, community mobilization, enabling local residents to assert their political voice over the issue of urban land allocation, finds its origins in just these national and local movements for decentralization. This period represented the beginning of democratization in Brazil, which is expressed, in the context of this chapter, as participation by the poor in the process of access to housing. It is within this context that the shift in housing policies from provider to stakeholder – which is effectively what the PREZEIS machinery entailed – could thus be construed as the result of both economic recession and local residents’ calls for affordable and effective housing provision. However, alarming numbers of squatter settlements continue to exist and grow in most of Brazil's cities. In terms of sheer quantity and scope, governmental housing programmes for the urban poor still occupy a distinct second place in comparison to private supply systems and the emergence of illegal settlements during recent decades.

Previous and current interventions have not, therefore, ameliorated the substandard housing conditions endured by most urban citizens. Rather, they have reinforced the notion that formal access to housing depends upon the inhabitants’ ability to pay. At the same time, since the private land and housing market has no self-regulatory device to ensure that prices remain affordable for all social groups, the poorest in Brazil (as in the rest of the developing world), must, instead, compete for space on illegally invaded land, sub-let rooms and occupy unused space on the streets.

The sustained and very large number of poor people living in urban areas in Brazil provides evidence that employment opportunities have not met the demand, nor has the housing system been able to provide affordable shelter for those households suffering from this imbalance.

Democratic structures, social welfare policies of the state and markets reveal a singularly incongruent relationship in Brazil. Nowhere is this more evident than in urban land and housing provision. While a combination of these three elements would appear to be a relevant amalgam for empowering the urban poor, historically, none has yet been successful in Brazil. Even in the current circumstances the combination of democratic reform (decentralization and new governance structures) and market enablement has not been successful in tackling the endemic conditions of the urban poor. Land tenure promoted during the time of the authoritarian welfare state was highly and unarguably unjust; and yet, paradoxically, the deficiency of land tenure security in the present climate is reminiscent of previous eras and regimes in significant ways.

Both the state and the market appear to neglect the demands of the poor for a better and more equitable quality of life. Sharing roles and responsibilities in the new neo-liberal models of local governance may be desirable ways of achieving better and more responsive living environments; but they may still be insufficient, in so far as the evidence from Brazil indicates, to improve the present state of affairs discussed above. Partnerships, flexibility, effectiveness and good governance are all buzzwords attracting attention from most sectors of society (both those in power and those under command). Thus, now and during the past, housing financiers, providers and developers (such as governments, private developers and international funding agencies) have helped the poor to house themselves by allocating resources to improve the supply of informal housing. These agencies have earmarked resources according to the ability of the poor to pay for land, building materials, technical support and services. However, where formal housing providers intervene, low-income groups become more economically, politically and technologically dependent upon housing bodies. Thus, the informal housing networks collapse. The poor, out of necessity rather than choice, thus continue their struggle to acquire housing through both legal and illegal self-help mechanisms.

These apparently contradictory paradigms appear to form a continuum in space and time where the prevailing elite benefit most. By neglect or by design, the poor and the needy have been excluded from the urban agenda. At best, the poor are on the social periphery, metaphorically and sometimes spatially, as well, suffering from the consequential insecurity and marginalization of state policies.

Brazil experienced 21 years of dictatorship, corresponding to the existence of the National Housing Bank. During this period, much public-sector housing was produced under the provider paradigm. But in Brazil, as in most countries subsequently undertaking the neo-liberal economic and governmental reform package, housing policies were strongly criticized for being inefficient and insufficient to tackle urban poverty and accommodate the huge number of people who needed housing most. This is certainly true in Brazil, and this chapter has presented the structural and procedural explanation for these outcomes.

But during the 16 years of democracy and state restructuring since that time, there is little, if any, evidence that a market economy has been any more successful in delivering land and housing which the poor can afford. Advocates of market enablement argue the capacity of markets to be flexible, efficient and more capable than the previous welfare structure to deliver basic social commodities such as housing. Yet, in Brazil, the markets are still far from responding to the ever-increasing demand for housing land by the urban poor. It could be argued, of course, that the reform package has not been fully elaborated. But, equally, it is also the case that the market itself is unable under any conditions to provide the poor with housing land, even where there is support from the state, as the cases of Recife, São Paulo, Belo Horizonte and others Brazilian cities suggest. Meanwhile, the informal and illegal settlements continue to proliferate at an unprecedented rate.

Accordingly, this chapter has argued that at the level of the individual household, neither state intervention in land and housing delivery, nor state promotion of market delivery processes have done much to reduce the problem of insecurity and marginalization of the urban poor. The failure of the state intervention model in Brazil to promote redistribution policies such as housing is attributable, at one level, to technical shortcomings, such as interest rate controls on housing lending, building standards and so on. At a structural level, this chapter has argued that the interests of the state during the military regime were played out in the co-option of the middle classes by producing an urban housing and industrialization nexus and de facto commodification of illegally settled urban land. With regard to the outcomes of a market enablement model, here the structural intentions of the policy agenda are explicitly directed to the marketing of critical urban resources, such as land and housing. Although there are fundamental structural contrasts between the two strategies, the ironic consequence is that they both essentially endorse or accede to market processes of land delivery that do little to minimize the huge gulf between the poorest and the richest. By analysing local processes of urban development ‘on the ground’ with respect to land tenure issues in Recife, the negative impacts of such policies for poor households have been demonstrated.

In reviewing the response of successive Brazilian governments to urban land and housing needs, the disjuncture between macro-economic policy and global trends, on the one hand, and local initiatives and outcomes to tackle urban poverty, on the other, could not be more striking. Social needs in terms of tenure security and housing supply for the urban poor remain vast. Neither the modalities of state intervention, nor the market have been able to narrow the huge gap in living conditions in Brazil's cities and to ameliorate the housing conditions of the urban poor in Brazil. Both strategies appear to intensify the risks of promoting social exclusion. Despite the rhetoric of promoting access to legal housing land for all, which is advocated as strongly in Brazil as elsewhere in the developing world, the outcomes remain illusory.

NOTES

1   The social interest of the BNH served only the middle class. The poor benefited very little from the fruits of the programme. In 1978, only 3 per cent of the subsidies were destined for the construction of houses for the poor (Bolaffi 1988).

2   See Souza (2000) for a comprehensive discussion on participatory budgeting in Brazil, and Tendler (1997) for the case of a successful democratic experience in Ceará state.

3   This chapter refers to ZEIS because it is the legal instrument used by the municipality of Recife to deal with informal city sectors. For a comprehensive discussion about ZEIS, see Maia (1995).

4   The ironic name of Iraque was used because the invasion occurred at the time of the first Gulf War.

5   Under the provisions of Brazilian law, usucapião (in effect, loss of compensatory entitlements) can be prescribed to privately owned urban land where possessors occupy plots that are up to 250 square metres for the purposes of housing themselves or their families, provided that they have been living in the plot for five consecutive years without opposition from the landowner.

REFERENCES

Alves, J F (1992) Metropoles: Cidadania e Qualidade de Vida, São Paulo, Editora Moderna

Andrade, L and Azevedo, S (1982) Habitação e Poder: Da Fundação da Casa Popular ao Banco Nacional da Habitação, Rio de Janeiro, Zahar Editores

Arretche, M (1990) ‘Intervenção do estado e setor privado: O modelo brasileiro de política habitacional’, Espaço e Debates, São Paulo, vol 31, pp21–36

Batley, R (1983) Power through Bureaucracy: Urban Political Analysis in Brazil, Aldershot, Gower

Bolaffi, G (1975) ‘Habitação e Urbanismo’, Ensaios de Opinião, vol 2, pp73–83

Bolaffi, G (1988) ‘Para uma nova política habitacional urbana: Possibilidades econômicas, alternativas operacionais e limites políticos’, in Valladares, L (ed) Habitação em Questão, Rio de Janeiro, Zahar Editores, pp167–196

Castro, M H G de (1991) ‘Descentralização e política social no Brasil: As perspectivas dos anos 90’, Espaços & Debates, vol XI, no 32, pp80–87

de Souza, F A M (1998) Perceived Security of Land Tenure and Low-income Housing Markets in Recife, Brazil, Unpublished PhD thesis, Oxford Brookes University

de Souza, F A M (1999a) ‘Land tenure security and housing improvements in Recife, Brazil’, Habitat International, vol 23, no 1, pp19–33

de Souza, F A M (1999b) ‘Sfidare i paradigmi della regolariizzazione fondiaria a Recife, Brasile’, Storia Urbana, no 23, July–December, pp119–145

de Souza, F A M (2004) ‘Security of land tenure revised: The case of CRRU in Recife and Porto Alegre, Brazil’, Habitat International, vol 28, pp231–244

FINEP–GAP (Finaciadora de Estudos e Projetos–Grupo de Arquitetura e Planejamento) (1983) Habitacao Popular: Inventario da Acao Governamental, Rio de Janeiro, Projeto

Klak, T (1990) ‘Spatially and socially progressive state policy and programs: The case of Brazil's national bank’, Annals of the Association of American Geographers, vol 80, pp571–589

Maia, M L (1995) ‘Land use regulations and rights to the city: Squatter settlements in Recife, Brazil’, Land Use Policy, vol 12, no 2, pp177–180

Maricato, E (1987) Politica Habitacional no Regime Militar: Do Milagre Brasileiro a Crise Economica, Petropolis, Vozes

Melo, M A B C de (1989) ‘Politicas publicas e transicao ddemocratica: Notas sobre o reformismo conservador na politica social e de habitacao’, in de Moura, A S (ed) O Estado e as Politicas Publicas na Transicao Democratica, São Paulo, Vertice–Massangana, pp166–188

Melo, M A B C de (1990) ‘Estruturação intra-urbana, regimes de acumulação e sistemas financeiros da habitação: Brasil em perspectiva comparada’, Espaço e Debates, São Paulo, vol 31, pp37–51

Melo, M A B C de (1996) ‘Crise federativa, guerra fiscal e “Hobbesianismo municipal”: Efeitos perversos da descentralização?’ São Paulo em Perspectiva, vol 13, no 10, pp11–20

Melo, M A B C de and de Moura, A S (1989) ‘Politicas publicas urbanas no Brasil: Uma analise dos projetos do Banco Mundial trabalho’, apresentado no XIII Encontro da ANPOCS, Caxambu 23–26 de Outubro de 1989, Texto para Discussao No 18, Recife, Universidade Federal de Pernambuco

Melo, M A B C, de Rezende, F and Lubambo, C (2000) Urban Governance, Accountability and Poverty: The Politics of Participatory Budgeting in Recife, Brazil, Report for the Project on Urban Governance, Partnership and Poverty, Birmingham, University of Birmingham, mimeo

Portes, A (1979) ‘Housing policy, urban poverty and the state’, Latin American Research Review, US, vol 14, no 2, p7

Sachs, C (1990) São Paulo: Politiques Publiques et Habitat Populaire, Paris, Editions de La Maison des Sciences de L'Homme

Shidlo, G (1990a) ‘Housing policy in Brazil’, in Shidlo, G (1990) (ed) Housing Policy in Developing Countries, London, Routledge

Shidlo, G (1990b) Social Policy in a Non-democratic Regime: The Case of Public Housing in Brazil, Oxford, Westview Press

Souza, C (2000) Participatory Budget in Brazilian Cities: Limits and Possibilities in Building Democratic Institutions, International Development Department of the School of Public Policy, Birmingham, University of Birmingham, mimeo

Tendler, J (1997) Good Government in the Tropics, Baltimore, John Hopkins University Press

Valenca, M M (1999a) ‘The closure of the Brazilian housing bank and beyond’, Urban Studies, vol 36, no10, pp1747–1768

Valenca, M M (1999b) ‘The lost decade and the Brazilian government's response in the 1990s’, The Journal of Development Areas, vol 33, pp1–52

Valladares, L (1988) ‘Urban sociology in Brazil: A research report’, International Journal of Urban and Regional Research, vol 12, pp285–303

World Bank (1989) Sites and Services and Low-cost Housing Project, Washington, DC, Project Performance Audit Report, World Bank