Disassembling connections: A comparative analysis of the politics of slum upgrading in eThekwini and São Paulo

Abstract

Chinese

This paper presents an innovative comparison that works creatively with the entangled spatialities of policy mobilities, drawing on a city-to-city cooperation between São Paulo (Brazil) and eThekwini (South Africa) municipalities for the exchange of slum upgrading expertise. The proposed comparative tactic entails tracing the establishment of this connection in order to disassemble the constituent flows and localities merged within it. Subsequently, by posing questions to one another, a relational comparison of the trajectory of slum upgrading policy in each locality is composed, unearthing the political and institutional conditions that preceded the existence of the connection per se. In that sense, both eThekwini and São Paulo are considered equivalent starting points from which local actors engaged in circulating ideas and mobilised slum upgrading policies. This paper not only brings a fresh approach to comparative methods – incorporating political contexts and their extensive overlapping networks of relations alongside a focus on particular policy trajectories – but also contributes to furthering global urban studies in two other ways. First, it provides insight into the processes by which policies are put on the move and localised (or not). Second, it demonstrates how repeated instances of urban practice may be unravelled by allowing each context of policy formation, with its distinctive trajectory of slum upgrading, to speak to one another. In this regard, the comparative analysis identified how, in both São Paulo and eThekwini, the consolidation of democracy was followed by the development of more technocratic approaches to the detriment of earlier slum upgrading initiatives focussed on community empowerment.

Keywords

comparative urbanism eThekwini policy mobilities São Paulo slum upgrading

Introduction

The British architect John FC Turner, after visiting Brazil in 1968, stated that what was presented to him as problems –favelas, mocambos, alagados,¹ he would consider as solutions. However, what was presented as solutions – low-cost housing estates – he called problems (Turner, 1968). Such ideas soon were appropriated by the World Bank, which began to finance infrastructure projects (sites and services and slum upgrading) in the so-called developing countries (Cohen, 2015). Three decades later, at the turn of the 21st century, the Cities Alliance was created by the World Bank and the United Nations, under the slogan of supporting the development of ‘Cities Without Slums’. Since then, this multi-donor coalition has been advocating worldwide for slum upgrading understood as a process through which informal areas are not only gradually improved through the provision of basic services, but land tenure and access to education and health care are also secured (Cities Alliance, n.d.).

The formation of the Cities Alliance, beyond reflecting growing concerns with the ‘continued proliferation of slums all over the world’ (World Bank, 2002: 20) can also be seen as a timely response to recognition of the increasingly prominent role of local governments and the interdependence between them, notably during the second United Nations Conference on Human Settlements held at Istanbul in 1996 (Borja and Castells, 1997). Shortly after the Cities Alliance was created, with a view to working closely with local governments, the United Cities and Local Governments (UCLG) was established, in 2004, as a global network of cities along with local, regional and metropolitan governments and their associations. Together, in 2009, these two organisations launched a Mentoring Programme to stimulate technical knowledge sharing amongst Southern cities (UCLG Peer Learning, 2013). Within this programme, a mentorship initiative on slum upgrading was agreed between the City Hall of São Paulo and the eThekwini Municipality² in 2011.

Despite the long international advocacy for slum upgrading and its wide-reaching repercussions in many urban contexts, there is a profound lack of comparative analysis of how this type of policy solution is produced in different cities. Only a few studies have analysed slum upgrading beyond the case of one single settlement or a few settlements within one city boundary (Huchzermeyer, 2004; Muchadenyika and Waiswa, 2018). While there has been some attention to slum upgrading as a circulating urban policy model on a global scale (Bertelli, 2021; Cochrane, 2011; Jones, 2012), the entanglements between developmental global circuits and local slum upgrading policies have not been sufficiently addressed.

Slum upgrading practices are, indeed, very localised as interventions are usually of great complexity due to intricate environmental and geotechnical issues, high population densities and social conflicts. Nevertheless, much could be gained by comparatively analysing slum upgrading in different localities. Bearing in mind the expanding policy activities of international organisations and transnational policy networks, linking local governmental agencies and professionals (Peck, 2003: 229), comparisons looking at distinctive but interconnected sites of policy development might be useful to unpack to what extent slum upgrading policies have been shaped by the circulating frameworks, or by the political and institutional specificities of each locality.

Moreover, composing a comparison of the distinct histories of slum upgrading policy in different contexts, intersected by transnational processes of slum upgrading policy dissemination, and observing the shared features and differentiated outcomes between them, could contribute to a collective effort to decentre the analytical perspective of urban studies (Robinson, 2006; Sheppard et al., 2013). Tracing slum upgrading policies draws attention to the urban experiences of poorer contexts, expanding the scope and themes of global urban studies. For instance, the examination of how ‘favelas’, ‘informal settlements’ and other urban forms resulting from peripheral or popular urbanisations (Caldeira, 2017; Streule et al., 2020) have been affected by governmental policies seeking to ‘upgrade’ these settlements, can extend conceptual insights on modes of making cities that start beyond the framing of official planning (Caldeira, 2017: 3).

Following these motivations, this paper focuses on the mentorship initiative on slum upgrading agreed between the City of São Paulo and the eThekwini Municipality. It proposes an innovative methodology to compare the slum upgrading policy trajectory of each locality, seeing these as interconnected and crosscut by a range of international circuits and organisations. In that sense, the empirical connection is understood not only as a conditioning context (Peck, 2011), or as a reason to explore the apparent tension between what is in motion and what is embedded in place (McCann and Ward, 2010) in relation to policy circulations. Rather, the policy connection is itself mobilised as a methodological tool to think comparatively about slum upgrading practices in São Paulo and eThekwini. Alongside their wider shared experiences of policy circulations, the long histories of slum upgrading policy in each context, assembled into a specific policy connection, have produced convergences and generated difference (Jacobs, 2006, 2012; Robinson, 2016, 2018). The approach proposed here treats the policy connection as a complex, assembled spatiality, which is then disassembled in analysis to generate comparative insights.

The paper first outlines this comparative method, and its theoretical and methodological underpinnings, before (re)tracing the connection as an empirical case to examine how the City Hall of São Paulo and the eThekwini Municipality, through their active participation in transnational events and networks, became linked. The comparison between São Paulo and eThekwini’s distinctive histories of slum upgrading policy is then composed through disassembling this link, highlighting the respective interplay of localised concerns and supralocal connections in key decision-making moments in each context. Finally, the historical political conditions underlying the partnership and shaping its impacts in each context are analytically revisited before the conclusions.

The analysis presented here is informed by research conducted between 2014 and 2019, which drew on an extensive exploration of the literature produced by Brazilian and South African scholars on the topic, supplemented by 20 in-depth interviews with policy actors from eThekwini and São Paulo and a critical examination of a variety of institutional materials (in different media formats) produced by international organisations, transnational municipal networks and municipalities. Participating in one of the official meetings of the partnership also allowed me a privileged opportunity to make observations.

This paper brings a fresh approach to comparative methods to extend global urban studies – it incorporates both the contextualised political arrangements and the networks of relations that intersect in, as well as connect and surpass different contexts, in the analysis of a singular policy connection. In disassembling this policy connection analytically, it brings into view the differentiated outcomes of circulating urban processes, as well as the distinctive trajectories of different urban contexts, for comparative reflection. The paper also contributes to furthering global urban studies in two other ways. First, it provides insight into the processes by which policies are put on the move and localised (or not). Second, it demonstrates how repeated instances of urban practice may be unravelled to allow each context of policy formation (with its trajectory of slum upgrading) to speak to one another and inspire further analysis. In this regard, the paper identifies that, in both São Paulo and eThekwini, the consolidation of democracy was followed by the development of more technocratic and top-down approaches to the detriment of earlier slum upgrading initiatives focussed on community empowerment.

Building an innovative comparative methodology

Since the 1970s cities have been often understood as central to capitalist development and comparatively analysed within a political economy paradigm. Largely influenced by dependency and world-systems theories (Cardoso and Faletto, 1970; Frank, 1967), many studies on Latin American and African cities (Abu-Lughod and Hay, 1977; Castells, 1973) interpreted some of the problems associated with them – rapid rural-urban migration, underemployment, regional imbalances and inadequate housing – as outcomes of their peripheral position in an unequal capitalist world-economy. Such a perspective chimes with what Tilly (1984: 83) has called the encompassing comparative method, whereby ‘different instances at various locations’ are taken as parts of the same system ‘on the way to explaining their characteristics as a function of their varying relations to the system as a whole’.

Although this encompassing method has enabled real advances in urban knowledge, especially in terms of recognising the interconnectedness among cities within the capitalist system (Robinson, 2011), it carries the risk of building a functionalist version of history, since a theoretical/logical whole is taken as a premise that determines the nature and functioning of individual units (McMichael, 1990; Tilly, 1984). Inspired by such systemic theories, for instance, ‘world cities’ (Friedmann, 1986) and ‘global cities’ (Sassen, 2000) approaches have often depicted cities at the periphery of the capitalist world-economy as dysfunctional and fundamentally different from those at the centre. In consequence, comparisons among cities in different hierarchical positions were discouraged, thus limiting the spectrum of cities informing urban theory (Robinson, 2006).

To overcome such methodological restraints and theoretical implications of encompassing comparisons a new comparative methodology is suggested here. It takes inspiration from the idea of relational comparisons proposed by Hart (2002, 2018) and incorporated comparisons found in McMichael (2000). These seek to avoid the adoption of an a priori theoretical and external whole as a frame and focus rather on a multiplicity of relations and connections shaping social processes. Theoretically, it draws on Massey’s (1994) relational conception of space and place (which also informed Hart’s relational comparisons), and on concepts of city and urban policy derived from it.

In Massey’s (1994) work, space-time is understood as ‘constructed out of the multiplicity of social relations across all spatial scales’ (p. 3), and place is conceived as a ‘subset’ of these multiple interactions or a ‘local articulation within a wider whole’ (p. 4). Following this conceptual framework, cities are seen as a ‘set of spaces where diverse ranges of relational webs coalesce, interconnect and fragment’ (Amin and Graham, 1997: 418). In consequence, urban policymaking is conceived as a relational phenomenon that goes beyond cities’ administrative boundaries (McCann and Ward, 2011; Peck and Theodore, 2015).

The comparative tactic proposed in this paper, therefore, takes seriously the recommendations to build experimental comparisons from such myriad connections (Robinson, 2011, 2018; Söderström, 2014 ; Ward, 2010). It makes use of a policy link between the City Hall of São Paulo and the eThekwini Municipality, through the UCLG’s Mentoring Programme, the Mentoring on Upgrading Informal Settlements, to build a comparison of the trajectories of slum upgrading policy in the two cities. The policy link is conceived as a spatiality, which gathers or assembles people, ideas and activities into proximity and closer relationships (Allen, 2008; McFarlane, 2011; Robinson, 2011). The link is assembled by actors from two different urban contexts, their historical and located experiences, as well as a number of international organisations and circuits, and widely circulating ideas, practices and documents.

Pragmatically, the proposed method entails tracing the constitution of this complex spatiality to subsequently disassemble it into its constituent parts (flows and localities), in order to historically examine the singular political and social conditions that made possible such a policy link, thereby shaping its impacts and outcomes. In other words, the cooperation established between São Paulo and eThekwini is methodologically used to unravel both the distinctive policy trajectories of the two cities and the larger historical dissemination of slum upgrading frameworks that intersect with and exceed these trajectories. For this reason, the comparative tactic was named ‘disassembling connections’.

The analysis builds on the vast urban policy mobilities literature (brilliantly compiled in Theodore [2019]) but both draws on and moves beyond the circulation itself to develop an explicit comparative method which brings the different contexts into mutual reflection (Robinson, 2018). In that sense, two territorialised instances of policy formation, wherein local actors are immersed in distinctive power relations and planning cultures (Robinson, 2013), are drawn into comparison. Consequently, in the ‘disassembling connections’ tactic, both localities are treated as equivalent starting points from which to understand how local actors engage in the circulating ideas of slum upgrading and mobilise policies.

Tracing the connection: Transnational circuits of slum upgrading policymaking

The story of the Mentoring on Upgrading Informal Settlements initiative (henceforth referred as ‘Mentoring’ initiative) can be traced back to June 2011, when eThekwini hosted a learning exchange meeting jointly organised by its Municipal Institute of Learning (MILE) and the Urban Strategic Planning Committee of UCLG. The general aim of this learning exchange event, focussing on governance and urban development, was to ‘promote the knowledge and capacity of cities in the global South’ through the comparison and evaluation of relevant practices, especially from Latin American cities (UCLG/MILE, 2011: 2). In practical terms, the event had the objective of establishing a new city-to-city cooperation through the then recently launched UCLG Mentoring Programme.

In the South African context, eThekwini has long been recognised as a source of local government expertise. The MILE, created in 2009, aimed to be Africa’s first learning institute for local government practitioners, focussing on activities to ‘re-capacitate municipal officials and empower them with a set of tools’ which would enable them ‘to meet more effectively their daily challenges’ (eThekwini Municipality, 2009: 1). Before the creation of MILE, senior managers of the eThekwini Municipality had been called upon to run training sessions for colleagues working in other South African municipalities, and throughout the Southern Africa Development Community region. The Institute was thus a response to what was identified as an ‘urgent need to co-ordinate and help streamline the various learning relationships that eThekwini has developed with a range of other municipalities, agencies and networks’ (eThekwini Municipality, 2009: 2).

The then eThekwini City Manager who envisaged the Institute, a former university professor and anti-apartheid activist, was at the time an influential member of the African National Congress (ANC), South Africa’s ruling party and the vice president of the UCLG Urban Strategic Planning Committee. His engagement with UCLG and efforts in promoting eThekwini as a learning city is one example – amongst several in this paper – of how localities are part of the assemblage of international circuits and connections.

The UCLG, created from a fusion of former associations – the International Union of Local Authorities (founded in 1913), the World Federation of United Towns and Cities (founded in 1957) and Metropolis (founded in 1984, still independently active) – can be seen as an expression of cities’ ambitions to make their unified voice heard by the UN and the World Bank (Saunier, 2008). This creation of a single partnership of municipal associations was, in turn, long awaited by these multilateral agencies. On the part of the World Bank, there was an ambition to move from providing direct technical assistance to national and subnational governments to a ‘wholesaling’ approach, that is, offering support to local government associations, research and training centres who would then take responsibility for the dissemination of knowledge and assistance once provided by the Bank (Kessides, 2000: 50, 62, 65–66).

The creation of the UCLG, as well as the Cities Alliance, were direct outcomes of the transformations in urban ‘development aid’ that emerged after Habitat II, with increasing emphasis given to the selection and promotion of ‘best practices’. This shift was linked to a discourse that learning amongst cities, through technical-assistance or capacity-building partnerships, would potentially improve urban governance, especially among cities in the South, thus increasing their competitiveness (Bontenbal, 2009; Peyroux, 2016). It was precisely this rationale that fostered the creation, in 2009, of a mentorship programme as part of the UCLG-Cities Alliance cooperation (Cities Alliance, 2011). Noteworthy for our analysis is that UCLG, mimicking the World Bank’s approach, ‘rather than a direct provider’ were conceived to be a ‘wholesale, learning institution, an anchor concerned with the overall role of continuous learning by cities and affiliated associations and organisations’ (Campbell, 2012: 2).

The 2011 event hosted in eThekwini was thus important as a trial for UCLG aspirations to sponsor peer to peer learning, as well as for MILE’s objectives to become a learning centre with good international visibility. On this occasion, Brazilian experiences of slum upgrading, local development and social participation were highlighted as ‘successful’ examples towards poverty reduction and the promotion of human rights, from which other cities could learn. Within the UCLG’s stated objective of fostering new mentoring projects, there was substantial motivation for eThekwini to explore learning from São Paulo. For MILE engaging in a mentorship with a renowned Southern American city could also enhance its prestige in the African context. In the words of one of the eThekwini’s practitioner, who co-organised the UCLG event:

It was a three-day meeting, one day we had to do (…) a matching service, like a dating service (…) in Africa we are always looking at South America, people used to look at the North, but now they are looking at South America as a similar context, a comparative context (…) [but] the only one partnership that Durban wanted to be engaged with was São Paulo slum upgrading. (…) Durban is seen as a well-resourced city in Southern African context or in an African context (…) But the one thing that did come through us was we want to learn about upgrading from São Paulo. (Former Head of Planning at the Municipality of eThekwini, 2015, personal communication)

Following this learning event, a three-and-a-half-year partnership was established between the municipality of eThekwini and the city of São Paulo – the ‘Mentoring’ initiative. The specific topics about which eThekwini wanted to learn from São Paulo were planning tools for the prioritisation and upgrading of settlements, strategies to acquire suitable land, financing methods and building standards, strategies for involving communities throughout the project cycle and job-creation initiatives (Municipal Institute of Learning [MILE], 2014). For São Paulo, according to the coordinator of the partnership for that city, this type of initiative with eThekwini at that time was of great interest, since the municipality had ‘highly qualified technicians’ (retired architect from the São Paulo Municipal Housing Secretariat, 2014, personal communication) capable of sharing the methodologies and models of upgrading projects that were then being implemented in São Paulo. There was thus a shared interest, as confirmed by the interviews undertaken for this research, between local practitioners of both cities and the UCLG staff in building a partnership.

Tracing a specific policy connection is a productive way to disentangle the many circuits of contemporary development aid and city-to-city cooperation, thus revealing the processes by which policies are put on the move and the role of organisations and key protagonists in such mobility (McCann, 2011; Wood, 2016). Furthermore, exploring this connection also sheds light on what Temenos and McCann (2012: 1391) have called the ‘local politics of policy mobility’, thus opening space to question the specificities of the slum upgrading policy trajectories in both cities collaborating in the cooperation. For instance, what were the political and institutional dynamics that allowed São Paulo to become a best practice case of slum upgrading, whereas eThekwini seemed to face constraints to implementing a citywide upgrading informal settlements programme, in such a way that actors there could envisage learning from São Paulo?

The next two sections explore the trajectory of slum upgrading policy in São Paulo and eThekwini to the moment of their assemblage into the ‘Mentoring’ initiative.

São Paulo: The making of a slum upgrading model with international mobility

When thinking about Brazil, certainly the most famous slum upgrading case that comes to mind is the ‘favela-bairro’ programme, started in Rio de Janeiro in the mid-1990s. São Paulo, however, also has its place in the slum upgrading hall of fame. In 2012, one year after the establishment of the learning partnership with the eThekwini Municipality, São Paulo’s Housing Secretariat (SEHAB, in Portuguese) was awarded the prestigious UN-Habitat Scroll of Honour for their work in this area.

The initial attempts to provide favelas with infrastructure in this city dates from 1979. It came in the context of the first step taken by the federal government towards the slum upgrading – the Promorar programme, which benefitted from the first World Bank’s loan to the country’s housing sector (World Bank, 1988). The Promorar programme – in conjunction with the Brazilian Urban Land Parcelling Law passed the same year – enabled the establishment of the first administrative structures aimed at upgrading favelas (see note 1) in many local governments. The removal of favelas was by then the dominant practice, despite some improved service provisions to favelas, in a context of increasing opposition from civic and religious organisations and social movements.

The first citywide programme to upgrade favelas, therefore, came only in 1989, during a municipal administration ruled by the then emergent left wing Workers’ Party (Bueno, 2000; Denaldi, 2003). This programme was developed in close dialogue with the housing grassroots movements (Bonduki, 2011) and drew largely on the upgrading expertise of other Brazilian cities, in terms of innovative techniques for providing infrastructure, respecting as much as possible a settlement’s original fabric, and creating instruments for land tenure security.³ Such slum upgrading projects were seen as an opportunity not only for housing repair, but also for social empowerment. They were marked by an ideal of knowledge co-production and often linked to residents’ self-management.

At the end of this progressive administration, in 1992, the first World Bank’s loan to the municipality was secured. This was part of a larger project, involving the state government and primarily aimed at cleaning up one of the city’s main water reservoirs, which was surrounded by favelas. The municipality’s role in this project was to upgrade these favelas. But the Guarapiranga programme, named after the reservoir, took place only in 1994, under the subsequent administration, which did not have the same grassroots links. The guiding principle was therefore no longer to work with slum dwellers but for them: priority was given to design, with public spaces and buildings for the relocation of families made by renowned architects.

It was from this point on that São Paulo’s favela upgrading experience began to be mobilised in international circuits. Soon after leaving work at the municipality in 2000, the former coordinator of the Guarapiranga programme joined the World Bank as a Senior Consultant. From this position, she took part in the Advisory Team that discussed a framework for upgrading Johannesburg’s Alexandra township – a site combining freehold plots, backyard shacks and free-standing informal settlements. The Guarapiranga programme itself was also widely disseminated internationally. For instance, it was spotlighted in an electronic manual on slum upgrading prepared by MIT (2001), at the request of the World Bank, to support the work of the Cities Alliance and integrated in the exhibition ‘Favelas Upgrading’ at the 2002 Venice Biennale of Architecture.

Because of São Paulo’s (and other Brazilian cities’) experience of slum upgrading, the Cities Alliance, shortly after its creation, considered Brazil to be a privileged site with which the institution should engage (Villarosa, 2016). This coincided with the advent of a new municipal administration in São Paulo which, again ruled by the Workers’ Party, sought the support of Cities Alliance to implement a citywide slum upgrading and land regularisation programme.

The so-called Bairro Legal programme received technical assistance and a small amount of financing from Cities Alliance (approximately US$300,000). This programme also largely benefitted from a national law that implemented the 1988 Federal Constitution principle of the social function of property. The Federal Law No. 10,257 of 2001, known as the City Statute, provided the necessary legal support to municipal governments to demarcate Special Zones of Social Interest (ZEIS), followed by approval of the Concession of Special Use for Housing Purposes (CUEM), which together allowed occupants of public land to claim their right to take advantage of it for housing purposes, provided that certain environmental and temporal criteria were respected. In the end, around 160 slum-occupied areas were reclassified with nearly 45,000 land tenure titles issued. Such results were widely broadcast, leading to the programme being recognised by international organisations as a best practice (Izar, 2004; World Bank, 2012). A key actor in this dissemination process was the former Head of Housing and Urban Development in São Paulo’s, who, after leaving the municipal administration, held strategic positions within the UN-HABITAT Task Force on Improving the Lives of Slum Dwellers (2003–2005) and joined the UN-HABITAT Advisory Group on Forced Evictions.

São Paulo’s image as a resourceful city of the South was also strengthened by the mayor of this last administration, who during her mandate played a prominent role in the negotiations for the creation of UCLG. As a result, she participated in the first co-presidency of the organisation between 2004 and 2007. When the next administration, in 2005, brought back the former Guarapiranga programme’s coordinator to lead the city’s housing policy, the making of São Paulo as a slum upgrading best practice took on a new chapter.

It is important to point out that slum upgrading was then strongly favoured by the National Growth Acceleration Programme (Programa de Aceleração do Crescimento – PAC, in Portuguese), launched by the federal government in 2007.⁴ São Paulo, where good expertise had been accumulated in the previous years, was one of the municipalities to receive the largest amounts of funding from the national programme. With substantial resources and renowned architecture firms being invited again to design the projects, the results of the citywide São Paulo Slum Upgrading Programme were meticulously publicised in several Metropolis and UCLG events (Madagascar in 2007, Mexico in 2009, Barcelona in 2010), in a good example of policy boosterism (McCann, 2013).

In 2008, with support from Cities Alliance, São Paulo hosted a South-South knowledge exchange on slum upgrading for representatives of Cairo, Manila, Lagos, Ekurhuleni and Mumbai. The remarkable quality of urban design and iconic architecture of São Paulo’s upgrading projects (Serapião, 2016) was also internationally promoted through international architecture journals, exhibitions and joint workshops with foreign universities on slum upgrading. It was this version of São Paulo slum upgrading policies that captured the eThekwini Municipality’s attention during the UCLG event held in that city in 2011.

EThekwini: Loss in situ upgrading expertise in favour of standardised housing delivery

In the South African case, in situ upgrading emerged in the 1980s. It was partly a strategy to minimise violent confrontations between dwellers of different political factions and the state, as well as between the factions (Hindson et al., 1993). Because of this violent context, participatory practices were challenging: ‘to get people together was a sticking point of the development’ (former Urban Foundation Consultant, 2015, personal communication). The first in situ upgrading experiments were implemented by the Urban Foundation (UF), a non-profit organisation created in 1976 by a group of South African businessmen. Their liberal view not only condemned the apartheid regime but was consistent with the World Bank’s urban development approach defending the end of state housing subsidies and promoting self-sustainable developments (Huchzermeyer, 2004; Smit, 1992). Besides the Urban Foundation, pioneering projects of in situ upgrading were also executed by service organisations voluntarily formed by leftist professionals and academics committed to fighting for the end of apartheid (Harrison et al., 2008). The most well-known programme in Durban was the Built Environment Support Group (BESG), founded in 1983. Initially their work was focussed on resisting forced removals with communities and advocating in situ upgrading as an alternative, strongly emphasising participatory forms of planning and decision-making, an aspect highlighted in interviews undertaken for this research. In the late 1980s, people from BESG visited São Paulo to create a link with experimental upgrading projects in the favelas (former member of BESG and former Director of Cato Manor Development Association, 2015, personal communication).

During the last decade of apartheid, therefore, the government’s response to squatter settlements was ambiguous: removals remained frequent, but interventions to improve these areas were also permitted (Harrison, 1992). In the early 1990s, two significant changes allowed upgrading experiences to scale up in Durban and South Africa at large: the creation of the Independent Development Trust, based on a donation made by the national government (Ameringen, 1995), and the subsequent approval of the first laws recognising the land rights of dwellers in established informal settlements (Harrison, 1992; Van Horen, 1996).

With the advent of the new democratic government in 1994, however, the national housing policy sidelined in situ upgrading practice in favour of the delivery of fully subsidised units produced by private developers for the poorest (Charlton and Klug, 2016). This policy model remained virtually unchanged for a decade, despite the burden it placed on the limited public coffers and the growing pressure for its revision (Huchzermeyer, 2004). Perhaps this was due to its strong link to the post-apartheid political settlement and its positive electoral effects for keeping the ANC in power (Huchzermeyer, 2011).

In Durban, however, where most of the in situ upgrading led by the Urban Foundation had occurred (Harrison, 1992), there was a local government attempt to sustain informal settlement upgrading via the national housing subsidy scheme, with more resources spent on infrastructure, rather than new houses and support given for self-construction. From 1997, with the establishment of a Metro Housing Unit, additional funding to top up the subsidies was put in place and numerous in situ upgrading projects were executed (Charlton, 2006; Charlton and Klug, 2016). Increasingly, however, these in situ upgrading projects drew on a rollover approach – that is, the complete demolition of previous built forms for land reparcelling and construction of new houses in one sector after another throughout the settlement. This change demonstrated the strength of the national model centred on the mass provision of free-standing houses (former official from the Durban Metro Housing Unit, 2015, personal communication).

The national government finally gave in to international and mainly domestic pressure to review the housing policy in the early 2000s. Specific issues concerned the persistent backlog of houses, the continuing presence of informal settlements and the costly and limited land suitable to develop, implying that new developments were far away (and often disconnected) from city’s amenities. The policy review was carried out by a team of local, provincial and national experts, with the technical assistance of the Cities Alliance. On this occasion, the South African Department of Housing came to Brazil to discuss its slum upgrading policy and visit projects (Huchzermeyer, 2006). Yet little attention was given to the methodologies that had been developed in South Africa in the 1990s (Charlton, 2006). The resulted policy amendment, approved in 2004, entitled ‘Breaking New Ground’, included a programme aimed at Upgrading Informal Settlements (UISP), envisioning infrastructure provision on the basis of limited relocations of families and participative processes.

Regardless of the UISP, the persistent inclination towards the delivery of subsidised houses, with strong support from the construction industry and significant electoral dividends (former member of BESG and former Director of Cato Manor Development Association, 2015, personal communication), ultimately distorted the meaning of in situ upgrading (Cirolia, 2014; Huchzermeyer, 2011). Instead, the new policy created a favourable scenario for the expansion of ‘rollover’ interventions and ‘greenfield’ developments, often in distant, poorly integrated areas and involving family relocations from better located informal settlements.

Local governments considered ‘rollover’ interventions to be less complex, and an easier way to comply with minimum standards for land titling, as well as cheaper and faster than in situ upgrading that sought to respect the original settlement’s layout and tried to keep housing demolition to a minimum (Charlton and Klug, 2016). In addition, ‘rollover’ interventions and ‘greenfields’ developments are often welcomed by informal settlement dwellers who see participation in upgrading activities as a burden, especially when they have the option to obtain a ready-built and free house from the government.⁵ There is also the lingering idea among dwellers that informal settlements are an expression of apartheid, which ended up being reinforced by the governmental discourse centred on the provision of new housing (Meth, 2020: 149, 151).

However, four years after the UISP’s launch none of the nine pilot upgrading projects announced had been implemented (Huchzermeyer, 2011), and the often violent eradication of informal settlements continued. The national government recognised that municipalities were in need of direct technical assistance as the ‘capacity and inclination to do upgrading in an incremental way involving communities is something that [South Africans] have somehow lost’ (Narsoo, cited in Charlton and Klug, 2016: 70).⁶ As result, the National Upgrading Support Programme (NUSP) was created in 2008, with the support of the World Bank and Cities Alliance (World Bank, 2016). Three years later, the World Bank brought officials from the city of São Paulo to present their experiences in South Africa (Huchzermeyer, 2011), confirming the interpretation that the capacity and even the inclination towards upgrading, incrementally and with the engagement of the communities, seemed to have disappeared in favour of state-led housing developments.

In parallel with the establishment of the NUSP, a technically oriented debate on intermediary housing solutions began to gain strength⁷ (Misselhorn, 2008). Pioneering the implementation of this approach nationally, the eThekwini Municipality launched the Interim Services Programme in 2008, to provide a range of incremental services – access roads, rainwater drainage, sanitation blocks (public bathrooms and showers), collective public taps and electricity distribution – and security of tenure for informal settlements considered to be suitable for full upgrading in the future (eThekwini Municipality, 2018).

Social movements’ collective resistance cannot be overlooked when analysing this shift in the municipality’s stance in relation to informal settlements (Ballard et al., 2006). For instance, the struggles of the grassroots movement, Abahlali baseMjondolo (‘people of the shacks’), against evictions resulted in an agreement with the eThekwini Municipality in 2009 for the development of 14 in situ upgrading projects. While the development of the interim services programme, combined with this move towards a more open dialogue, is appreciated by the movement, it has, however, continued to call for greater participation in decision-making processes and to denounce the violence of forced evictions that continued to occur (founding member and former president of the Abahlali baseMjondolo, 2015, personal communication).

‘Joining the ends’: Reassembling the connection to compare policy histories

In São Paulo, pioneer slum upgrading initiatives were developed in the context of the country’s return to democracy in the late 1980s. These were marked by a strong sense of popular participation and civic organisations’ engagement around a project of urban reform. During the 1990s, slum upgrading increasingly scaled-up and the support of international organisations was quite significant in this process. The consequences of this expansion were that communities lost much of their past protagonist role in decision-making while São Paulo’s international visibility within (and influence on) slum upgrading circuits became increasingly prominent. Through close interaction with the World Bank, Cities Alliance and UCLG, different political actors from São Paulo acted as ‘transfer entrepreneurs’ (Stone, 2004: 549) engaged in ‘transnational policy communities’ of experts and professionals, sharing their expertise and information to form ‘common patterns of understanding regarding policy through regular interaction’ (Stone, 2004: 559).

One line of connection fostered through São Paulo’s solid collaboration with international organisations and transnational municipal networks was that with the eThekwini Municipality. There, in situ upgrading had played a significant role in the waning years of apartheid and the early years of democracy but was then passed over for a model of housing provision for the poorest based on fully-subsidised and free-standing houses produced by private developers. The centrality given by national government to this model was a key factor in the failure of Durban’s attempt to improve and scale up comprehensive in situ upgrading in the early 2000s. In consequence, a gap emerged in the expertise within the municipality, which soon became a further constraint on implementing in situ upgrading. Added to this was the dream nurtured by many poor families of owning a standard plot with a ready-built house, a dream that has been fed back to influence the perspective of political candidates in each election.

Within this context, the ‘Mentoring’ initiative between the City of São Paulo and the eThekwini Municipality was one of many South African approaches to Brazilian authorities to learn from their slum upgrading experience. When the then City Manager was asked to comment on the interest in São Paulo’s slum upgrading model in the light of the past recognised in situ upgrading experience in Durban, he emphasised that São Paulo’s approach at that time was more ‘design-led’, whereas Durban’s experience had been more ‘planning-led’ (former eThekwini City Manager, 2015, personal communication. Of all the participants involved in the learning partnership, as a former BESG member he was the only one who had participated in Durban’s pioneer in situ upgrading projects.) eThekwini’s report on the results of the learning initiative indeed pointed out a need to look beyond the provision of core housing units and road infrastructure in upgrading and greenfield projects (MILE, 2014: 7), recommending special attention to the design of urban form through higher density housing units, public spaces and facilities, as well as environmentally friendly alternatives, such as creating a linear park which they implemented in the pilot area of Welbedagt.

However, another highlighted lesson was the importance of social surveys and community participation in upgrading interventions, something that, according to one of the interviewees, ‘in South Africa [was] never heard of, never done’ (former Head of Planning at the Municipality of eThekwini, 2015, personal communication). A statement that certainly appears to be at odds with the trajectory of in situ upgrading in Durban examined in this paper, especially its pioneer attempts which were strongly participatory.

Perhaps the most important outcome of the ‘Mentoring’ initiative was that the lessons learned by the eThekwini Municipality – presented as a well-resourced and performing city ‘still learning from another city’– could then be ‘shared with other South African cities as well as national and provincial departments’ (MILE, 2014: 13). Following the line of thought that ‘the city that learns is also the city that teaches’, the link with São Paulo seems to correspond to MILE’s expectations of enhancing its expertise and image as a learning centre. Moreover, it matched not only UCLG’s aspirations of up-scaling learning experiences internationally, confirming its intermediary role in brokering connections, but also supported the Cities Alliance’s agenda to promote slum upgrading and the World Bank’s move towards a ‘wholesaling’ approach in technical assistance.

Harrison (2015: 206) pointed out, in his analysis of some cooperation links established by Johannesburg with cities in the global South, including in Brazil, that there are often ‘inevitable inequalities in power, contextual differences and complex political agendas at play’ in technical knowledge sharing. Within the complex spatiality of a policy link as assemblage, these power relations and interests shape the ongoing policy circulations as much as the outcomes in localities.

Conclusion

This paper draws creatively on a slum upgrading policy connection to suggest an innovative comparative practice, in which linked localities (São Paulo and eThekwini, in this case) are treated as equivalent starting points from which local actors engage in and shape slum upgrading policy circuits. The disassembling connections’ tactic allows rethinking the making of cities in their relationality, through the long-standing circuits of urban development aid, policy circulations and transnational city networks. These cities and their distinctive trajectories of slum upgrading policies are brought into view through the examination of the policy link that put them together.

Tracing the policy connection between São Paulo and eThekwini disclosed the meanders of slum upgrading policy mobility: the entanglements between global circuits of urban development policy and local processes, as well as the many power-laden relations implicated in them. This paper subsequently composed a comparison of the evolution of policies towards upgrading informal settlements in each city, through the lens of questions posed by the policy connection. This revealed further contextual issues and helped to identify the sequence of events which explains the development of the city-to-city cooperation. Therefore, the historical trajectories of slum upgrading policies in eThekwini and São Paulo presented here can be considered as resulting from following (one of) their shared policy connection(s) and excavating some of the local processes, institutional dynamics and political actors that have contributed to shaping it.

Furthermore, in allowing one case to speak to another, the ‘disassembling connections’ method permits us to see policy outcomes as repeated instances produced under conditions of difference. It brings into view an ironic relationship between the deepening of democracy and the ‘depoliticisation’ of slum upgrading projects. In other words, it unveiled how slum upgrading initiatives, initially based on fostering community empowerment, lost much of their transformative potential pari passu with the consolidation of democracy. In São Paulo, this occurred through the combination of advanced technical solutions with formal and restrained participatory methodologies, while in eThekwini this was the result of the implementation of a formal subsidised housing programme and, more recently, a focus on basic services provision. In both contexts, therefore, there was a shift towards greater conservatism in slum upgrading practices with democratisation. Of course, the dynamic politics underlying the slum upgrading histories in the two contexts and their shared wider policy circuits, mean that this should not be seen as a static or permanent outcome.

Footnotes

Acknowledgements

The PhD research from which this paper draws was supervised by Professor Carlos Vainer. It also benefitted enormously from the guidance of Professor Jennifer Robinson, during my visiting research period at the University College London. I am deeply grateful to them as well as to all the interviewees in Brazil and South Africa. Some initial insights were presented on the 2015 American Association of Geographers Annual Meeting, in Chicago, and the 2018 International Conference on Policy Diffusion and Development Cooperation, in São Paulo. I thank the helpful comments received on these occasions, especially from Astrid Wood. I also thank the constructive suggestions received from the editor of this special issue Jenny Robinson and the three anonymous reviewers on earlier versions of the manuscript. Responsibility for the final content remains my own.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by FAPERJ Grade 10 Scholarship (E-26/100.241/2014), CNPq SWE Scholarship (248950/2013-4) and IJURR Foundation Writing Up Grant (2019). This paper was written whilst supported by Urban Studies Foundation as Postdoctoral Research Fellow.

ORCID iD

Camila Saraiva

Notes

References

Abu-Lughod

Hay

(1977) Third World Urbanization. Chicago, IL: Maaroufa Press.

Allen

(2008) Powerful geographies: Spatial shifts in the architecture of globalization. In: Clegg

Haugaard

(eds) Handbook of Power. London: SAGE, pp.157–174.

Ameringen

(1995) Building a new South Africa. In: Urban Policy, vol. 2. Ottawa: IDRC.

Amin

Graham

(1997) The ordinary city. Transactions of the Institute of British Geographers 22(4): 411–429.

Ballard

Habib

Valodia

(eds) (2006) Voices of Protest: Social Movements in Post-Apartheid South Africa. Scottsville, VA: University of KwaZulu Press.

Bertelli

(2021) What kind of global city? Circulating policies for ‘slum’ upgrading in the making of world-class Buenos Aires. Environment and Planning A: Economy and Space 53(6): 1293–1313.

Bonduki

(2011) Entre a academia e a política: uma trajetória entrelaçada. Habilitation Thesis, Universidade de São Paulo, Brazil.

Bontenbal

(2009) Strengthening urban governance in the south through city-to-city cooperation: Towards an analytical framework. Habitat International 33: 181–189.

Borja

Castells

(1997) Local & Global: Management of Cities in the Information Age. London: UNCHS/Earthscan Publications.

10.

Bueno

LMM

(2000) Projeto e favela: metodologia para projetos de urbanização. PhD Thesis, Universidade de São Paulo, Brazil.

11.

Caldeira

(2017) Peripheral urbanization: Autoconstruction, transversal logics, and politics in cities of the global south. Environment and Planning D Society and Space 35(1): 3–20.

12.

Campbell

(ed.) (2012) Outlook for learning: Strategic note for UCLG and metropolis. Available at: https://www.uclg-cisdp.org/en/outlook-city-learning-strategic-note-united-cities-and-local-governments-and-metropolis (accessed 7 March 2018).

13.

Cardoso

Faletto

(1970) Dependência e Desenvolvimento na América Latina: Ensaios de Interpretação Sociológica. São Paulo: Editora Afiliada.

14.

Castells

(1973) Imperialismo y urbanización en América Latina. Revista Planificación 8: 1–18.

15.

Charlton

(2006) Learning from the local: Experiences of informal settlement upgrading in Kwazulu-Natal. South African Review of Sociology 37(1): 48–64.

16.

Charlton

Klug

(2016) Pressures on practice: How RDP housing and other factors have shaped informal settlement upgrading in eThekwini municipality. In: Cirolia

(ed.) Upgrading Informal Settlements in South Africa. Cape Town: UCT Press, pp.56–76.

17.

Cirolia

(2014) South Africa’s emergency housing programme: A prism of urban contest. Development Southern Africa 31(3): 397–411.

18.

Cities Alliance (2011) UCLG and cities alliance: Working together to support strategic urban planning. Available at: https://www.citiesalliance.org/newsroom/news/spotlight/uclg-and-cities-alliance-working-together-support-strategic-urban-planning (accessed 8 March 2018).

19.

Cities Alliance (n.d.) Slums and slum upgrading. Available at: http://www.citiesalliance.org/About-slum-upgrading#Why_is_slum_upgrading_important (accessed 8 March 2018).

20.

Cochrane

(2011) Making up global urban policies. In: Bridge

Watson

(eds) The New Blackwell Companion to the City. Chichester: Wiley-Blackwell, pp.738–746.

21.

Cohen

(2015) John FC turner and housing as a verb. Built Environment 41(3): 412–418.

22.

Denaldi

(2003) Políticas de urbanização de favelas: evolução e impasses. PhD Thesis, Universidade de São Paulo, Brazil.

23.

eThekwini Municipality (2009) Rationale for a municipal institute of learning – MILE: Enhancing local government capacity: For practitioners, by practitioners. Available at: http://www.mile.org.za/QuickLinks/Documents/MILE_Founding%20Document.pdf (accessed 13 April 2018).

24.

eThekwini Municipality (2018) Strategy, principles and approaches to the informal settlement upgrading programme: Lessons from eThekwini municipality. Available at: http://www.mile.org.za/AFUS2018/AFUS%202018%20Presentations/Day%201-Strategy-Principles%20and%20approaches.pdf (accessed 25 June 2019).

25.

Frank

(1967) Capitalism and Underdevelopment in Latin America: Historical Studies of Chile and Brazil. New York, NY: Monthly Review Press.

26.

Friedmann

(1986) The world city hypothesis. Development and Change 17: 69–83.

27.

Harrison

(1992) The policies and politics of informal settlement in South Africa: A historical perspective. Africa Insight 22(1): 14–22.

28.

Harrison

(2015) South–south relationships and the transfer of ‘best practice’: The case of Johannesburg, South Africa. International Development Planning Review 37(2): 205–223.

29.

Harrison

Todes

Watson

(2008) Planning and Transformation: Learning From the Post-Apartheid Experience. London: Routledge.

30.

Hart

(2002) Disabling Globalization: Places of Power in Post-Apartheid South Africa. Berkeley, CA: University of California Press.

31.

Hart

(2018) Relational comparison revisited: Marxist postcolonial geographies in practice. Progress in Human Geography 42(3): 371–394.

32.

Hindson

Byerley

Morris

(1993) From violence to reconstruction: The making, disintegration and remaking of an apartheid city. CSDS working paper, 10, pp.1–36. Durban: University of Natal.

33.

Huchzermeyer

(2004) Unlawful Occupation Informal Settlements and Urban Policy in South Africa and Brazil. Trenton, NJ: Africa World Press.

34.

Huchzermeyer

(2006) The new instrument for upgrading informal settlements in South Africa: Contributions and constraints. In: Huchzermeyer

Karam

(eds) Informal Settlements: A Perpetual Challenge?Cape Town: UCT Press, pp.41–61.

35.

Huchzermeyer

(2011) Cities With ‘Slums’: From Informal Settlement Eradication to a Right to the City in Africa. Claremont: UCT Press.

36.

Izar

(2004) Integrating the Poor: Urban Upgrading and Land Tenure Regularization in the City of Sao Paulo. Washington, DC: World Bank.

37.

Jacobs

(2006) A geography of big things. Cultural Geographies 13(1): 1–27.

38.

Jacobs

(2012) Commentary – Comparing comparative urbanisms. Urban Geography 33(6): 904–914.

39.

Jones

(2012) ‘Bankable Slums’: The global politics of slum upgrading. Third World Quarterly 33(5): 769–789.

40.

Kessides

(2000) Cities in Transition: World Bank Urban and Local Government Strategy. Washington, DC: World Bank.

41.

McCann

(2011) Urban policy mobilities and global circuits of knowledge: Toward a research agenda. Annals of the Association of American Geographers 101(1): 107–130.

42.

McCann

(2013) Policy boosterism, policy mobilities, and the extrospective city. Urban Geography 34(1): 5–29.

43.

McCann

Ward

(2010) Relationality/territoriality: Toward a conceptualization of cities in the world. Geoforum 41(2): 175–184.

44.

McCann

Ward

(eds) (2011) Mobile Urbanism: Cities and Policymaking in the Global Age. Minneapolis, MN: University of Minnesota Press.

45.

McFarlane

(2011) The city as assemblage: Dwelling and urban space. Environment and Planning D Society and Space 29(4): 649–671.

46.

McMichael

(1990) Incorporating comparison within a world-historical perspective: An alternative comparative method. American Sociological Review 55(3): 385–397.

47.

McMichael

(2000) World-systems analysis, globalization, and incorporated comparison. Journal of World-Systems Research 6(3): 668–689.

48.

Massey

(1994) Space, Place, Gender. Minneapolis, MN: University of Minnesota Press.

49.

Meth

(2020) ‘Marginalised formalisation’: An analysis of the in/formal binary through shifting policy and everyday experiences of ‘poor’ housing in South Africa. International Development Planning Review 42(2): 139–164.

50.

Misselhorn

(2008) Position paper on informal settlements upgrading: Part of a strategy for the second economy. Report for urban LandMark, Pretoria, South Africa, April.

51.

MIT (2001) Upgrading urban communities – A resource framework. Available at: http://web.mit.edu/urbanupgrading/upgrading/whatis/what-is.html (accessed 27 April 2015).

52.

Muchadenyika

Waiswa

(2018) Policy, politics and leadership in slum upgrading: A comparative analysis of Harare and Kampala. Cities 82: 58–67.

53.

Municipal Institute of Learning (MILE) (2014) Mentoring on upgrading informal settlements. Report, Metropolis Initiatives, Durban, South Africa, July.

54.

Peck

(2003) Geography and public policy: Mapping the penal state. Progress in Human Geography 27(2): 222–232.

55.

Peck

(2011) Geographies of policy: From transfer-diffusion to mobility-mutation. Progress in Human Geography 35(6): 773–797.

56.

Peck

Theodore

(2015) Fast Policy: Experimental Statecraft at the Thresholds of Neoliberalism. Minneapolis, MN: University of Minnesota Press.

57.

Peyroux

(2016) Circulation des politiques urbaines et internationalisation des villes: la stratégie des relations internationales de Johannesburg. EchoGéo 36: 1–18.

58.

Robinson

(2006) Ordinary Cities: Between Modernity and Development. London: Routledge.

59.

Robinson

(2011) Cities in a world of cities: The comparative gesture. International Journal of Urban and Regional Research 35(1): 1–23.

60.

Robinson

(2013) “Arriving at” urban policies/the urban: Traces of elsewhere in making city futures. In: Söderström

(ed.) Critical Mobilities. New York, NY: Routledge, pp.1–28.

61.

Robinson

(2016) Thinking cities through elsewhere: Comparative tactics for a more global urban studies. Progress in Human Geography 40(1): 3–29.

62.

Robinson

(2018) Policy mobilities as comparison: Urbanization processes, repeated instances, topologies. Revista de Administração Pública 52(2): 221–243.

63.

Sassen

(2000) Cities in a World Economy. London: Pine Forge Press.

64.

Saunier

P-Y

(2008) Introduction: Global city, take 2: A view from urban history. In: Ewen

Saunier

P-Y

(eds) Another Global City: Historical Explorations Into the Transnational Municipal Moment 1850–2000. New York, NY: Palgrave, pp.1–18.

65.

Serapião

(2016) Linking the formal and informal: Favela urbanisation and social housing in São Paulo. Architectural Design 86(3): 70–79.

66.

Sheppard

Leitner

Maringanti

(2013) Provincializing global urbanism: A manifesto. Urban Geography 34(7): 893–900.

67.

Smit

(1992) Urban foundation: Transformation possibilities. Transformation 18: 35–42.

68.

Söderström

(2014) Cities in Relations: Trajectories of Urban Development in Hanoi and Ougadougou. Oxford: Wiley-Blackwell.

69.

Stone

(2004) Transfer agents and global networks in the ‘Transnationalization’ of policy. Journal of European Public Policy 11(3): 545–566.

70.

Streule

Karaman

Sawyer

, et al. (2020) Popular urbanization: Conceptualizing urbanization processes beyond informality. International Journal of Urban and Regional Research 44(4): 652–672.

71.

Temenos

McCann

(2012) The local politics of policy mobility: Learning, persuasion, and the production of a municipal sustainability fix. Environment and Planning A: Economy and Space 44(6): 1389–1406.

72.

Theodore

(2019) Policy mobilities. In: Warf

(ed.) Oxford Bibliographies in Geography. New York, NY: Oxford University Press. DOI: 10.1093/obo/9780199874002-0205.

73.

Tilly

(1984) Big Structures, Large Processes, Huge Comparisons. New York, NY: Russell SAGE Foundation.

74.

Turner

JFC

(1968) Habitação de Baixa Renda no Brasil: Políticas atuais e oportunidades futuras. Revista Arquitetura IAB 68: 17–19.

75.

UCLG/MILE (2011) Peer learning exchange: Governance and urban development. Report, Durban, South Africa, June.

76.

UCLG Peer Learning (2013) 04 Mentoring story: Lilongwe & Johannesburg experience (2008–2012). Available at: http://issuu.com/uclgcglu/docs/jb_lilongwestory?e=5168798/1337741 (accessed 8 March 2018).

77.

Van Horen

(1996) Informal settlement upgrading in Durban, South Africa: Building institutional capacity to sustain the improvement process. PhD Thesis, University of British Columbia, Canada.

78.

Villarosa

(2016) Impacto da Cities Alliance no Brasil: uma análise crítica. Report, Cities Alliance, 5 December 2016 (manuscript), p.78.

79.

Ward

(2010) Towards a relational comparative approach to the study of cities. Progress in Human Geography 34(4): 471–487.

80.

Wood

(2016) Tracing policy movements: Methods for studying learning and policy circulation. Environment and Planning A: Economy and Space 48(2): 391–406.

81.

World Bank (1988) Brazil: Sites and services and low-cost housing project. Report no. 7581. Washington, DC: World bank.

82.

World Bank (2002) Cities Alliance Annual Report 2002. Washington, DC: World Bank.

83.

World Bank (2012) Promoting Innovation in Urban Upgrading in Sao Paulo. Cities Alliance in Action. Washington, DC: World Bank Group.

84.

World Bank (2016) The national upgrading support program: Securing South Africa’s urban future. Available at: https://www.worldbank.org/en/news/feature/2011/12/16/the-national-upgrading-support-program-securing-south-africas-urban-future (accessed 2 March 2021).