Vol. 30, n. 2, jul/dez, 2024
ISSN: 2179-6807 (online)
123
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
COMPETIÇÃO E COOPERAÇÃO NA CENA URBANA: SEGREGAÇÃO RACIAL
NA REGIÃO METROPOLITANA DE SÃO PAULO
Burkay Koseoglu
1
Resumo: O estudo examina a atual segregação socioespacial na região metropolitana de São
Paulo. Usando ferramentas quantitativas como Summary Statistics, Location Quotient (LQ),
Global Moran's I, Local Indicators of Spatial Association (LISA) e agrupamento K-Means na
escala municipal, ele revela como os limites entre grupos raciais são moldados por condições
históricas, políticas, sociais, culturais e econômicas. Os resultados indicam padrões espaciais
distintos. Enquanto as populações indígenas estão concentradas em áreas rurais e semi-rurais,
as populações pretas e pardas apresentam níveis moderados a altos de agrupamento nos
municípios vizinhos, relacionados à segregação histórica e à desigualdade sistêmica. Além
disso, as populações amarela e especialmente as brancas estão mais integradas em áreas
economicamente desenvolvidas. A análise também destaca a inter-relação de fatores
socioeconômicos. As descobertas contribuem para a compreensão atual de como a dinâmica
social molda o acesso dos grupos raciais a recursos e oportunidades potenciais.
Palavras-chave: Competição social. Cooperação social. Identidade social. Segregação
socioespacial. Grande São Paulo
COMPETITION AND COOPERATION ON THE URBAN SCENE: RACIAL SEGREGATION IN
METROPOLITAN SÃO PAULO
Abstract: The study examines the current sociospatial segregation in Metropolitan São Paulo.
Using quantitative tools such as the Summary Statistics, Location Quotient (LQ), Global
Moran's I, Local Indicators of Spatial Association (LISA) and K-Means clustering at the
municipality scale, it reveals how the boundaries between racial groups are shaped by
historical, political, social, cultural and economic conditions. The results indicate distinct spatial
patterns. While the indigenous populations are concentrated in rural and semi-rural areas, the
black and brown populations show moderate to high levels of clustering in the surrounding
municipalities, related to historical segregation and systemic inequality. In addition, the yellow
and especially the white populations are more integrated in economically developed areas.
The analysis also highlights the interrelationship of socioeconomic factors. The findings
1
Planejador Urbano e Regional. Masterado pela Universidade de Gazi, Ancara/TURQUIA. Doutorando
em Programa de Pós-Graduação em Arquitetura e Urbanismo da Faculdade de Arquitetura e Urbanismo
da Universidade de Brasilia (PPG-FAU-UNB). ORCID iD: https://orcid.org/0009-0007-3215-0985. E-mail:
burkaykoseoglu@gmail.com.
https://www.periodicos.unimontes.br/index.php/rds/
124
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
contribute to the current understanding of how social dynamics shape racial groups’ access to
resources and potential opportunities.
Keywords: Social competition. Social cooperation. Social identity. Socio-spatial segregation.
Metropolitan São Paulo
COMPETENCIA Y COOPERACIÓN EN LA ESCENA URBANA: SEGREGACIÓN RACIAL EN EL
METROPOLITANO DE SÃO PAULO
Resumen: El estudio examina la segregación socioespacial actual en el área metropolitana de
São Paulo. Mediante el uso de herramientas cuantitativas como las Estadísticas de Resumen, el
Cociente de Localización (LQ), el Índice de Moran Global, los Indicadores Locales de Asociación
Espacial (LISA) y la agrupación K-Means a escala municipal, se revela cómo las fronteras entre
los grupos raciales están determinadas por las condiciones históricas, políticas, sociales,
culturales y económicas. Los resultados indican patrones espaciales distintos. Mientras que las
poblaciones indígenas se concentran en áreas rurales y semirrurales, las poblaciones negras y
pardas muestran niveles moderados a altos de agrupación en los municipios circundantes,
relacionados con la segregación histórica y la desigualdad sistémica. Además, las poblaciones
amarillas y especialmente las blancas están más integradas en áreas económicamente
desarrolladas. El análisis también destaca la interrelación de factores socioeconómicos. Los
hallazgos contribuyen a la comprensión actual de cómo la dinámica social determina el acceso
de los grupos raciales a los recursos y las oportunidades potenciales.
Palabras-clave: Competición social. Cooperación social. Identidad social. Segregación
socioespacial. Área metropolitana de São Paulo
INTRODUÇÃO
Cities are not only physical arrangements but also social structures defined by
power relations, historical processes and identity constructions. Thus, the socio-spatial
segregation is the outcome of economic, political, social and cultural dynamics
(BOURDIEU, 1986; LEFEBVRE, 1991; QUIJANO, 2000). Within these social structures, in
addition to material gains such as the acquisition of urban resources, abstract gains
such as identity and social acceptance lead to the emergence of competition and
cooperation strategies among individuals and social groups. In this process, in addition
to many identity elements, racial identities may become determining factors in the
amount of urban resources acquired and in the struggle to find a place in physical
space. As a result, social competition and cooperation shape power struggles and
resource sharing among individuals and groups, while also playing an important role in
the production of segregated spaces as an outcome (TAJFEL; TURNER, 1979; TURNER
et al., 1987).
https://www.periodicos.unimontes.br/index.php/rds/
125
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Henri Lefebvre’s theory of the social production of space provides a framework
for understanding this issue. The theory emphasizes the partiality and dynamism of
urban spaces by distinguishing between perceived space, designed space and lived
space. Thus, urban space is constantly reproduced through social interactions, power
struggles and ideological practices. As a result, discrimination appears as a dynamic
process rather than a fixed outcome (LEFEBVRE, 1991). Edward Soja, drawing on
Lefebvre, connects the mental and social dimensions of space with the concept of the
third space and emphasizes the importance of spatial justice in addressing inequalities
(SOJA, 2000). On the other hand, Pierre Bourdieu explains the mechanisms of socio-
spatial segregation through the concepts of habitus and capital, where cultural capital
maintains access to privileged urban spaces and excludes disadvantaged groups
(BOURDIEU, 1986; BOURDIEU, 1990). Loic Wacquant completes this perspective by
examining spatial exclusion and stigma, showing how urban spaces are used as
mechanisms of social exclusion and control (WACQUANT, 2008).
From an anti-colonialist perspective, Aníbal Quijano’s concept of coloniality
(QUIJANO, 2000) and Silvia Rivera Cusicanqui’s ch’ixi metaphor (CUSICANQUI, 2012)
highlight the enduring legacy of colonial hierarchies and the conflict between
indigenous and marginalized groups. Walter Mignolo and Boaventura de Sousa Santos
also expand the epistemic dimension by criticizing the dominance of Western-centric
knowledge systems that perpetuate inequalities (MIGNOLO, 2011; SANTOS, 2014). In
addition, Frantz Fanon underlines the psychological dimension by showing how
colonial systems impose new identities that reinforce urban racial and cultural
hierarchies (FANON, 1961). Focusing on racism, George Lipsitz explains how urban
planning practices institutionalizes discrimination through exclusionary practices
(LIPSITZ, 2006).
Furthermore, David Harvey and Saskia Sassen offer insights into the capitalist
and globalized nature of urban segregation. They explain, on the one hand,
segregation is associated with class inequalities and collective consumption (HARVEY,
2009), while on the other hand, global forces reshape urban areas and intensify socio-
spatial divisions (SASSEN, 2001). When these theories are combined, it is understood
that urban segregation is not simply the result of local factors, but a complex and
dynamic product of intersecting global, historical and social processes.
https://www.periodicos.unimontes.br/index.php/rds/
126
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
It should be emphasized that social identities, which emerge depending on
context, individual perceptions and group dynamics, are not fixed. Rapid and abundant
social interactions and complex spatial arrangements accelerate this dynamism
(TAJFEL; TURNER, 1979; TURNER et al., 1987). Considerable difficulties will be
encountered when the issue is addressed on an urban scale due to its dynamic and
multifaceted nature. However, the dynamics of social competition and cooperation
provide an important perspective for understanding human behavior in urban
environments.
Metropolitan São Paulo, with its colonial legacy, social structures founded on
this legacy, and rapidly growing population due to local as well as global influences, is a
suitable example for addressing the relationships between social groups shaped by
different dominant identity elements, together with the historical and structural
mechanisms that determine these relationships. Its current situation has been shaped
by intense capitalist and global expansion processes on the one hand, and by
segregationist policies that reinforce racial and class hierarchies on a spatial level on
the other. Thus, it is important in terms of showing how the results of relations
between social groups are embodied in space.
This study analyzes the spatial distribution of the racial categories defined by
the Brazilian Institute of Geography and Statistics (IBGE) such as indigenous, black,
brown, yellow and white population groups. The study examines the settlement
patterns of these groups within the framework of social competition and cooperation
dynamics and aims to present the current results in light of the causes of socio-spatial
segregation in the process. For this purpose, quantitative spatial analysis methods such
as Summary Statistics, Location Quotient (LQ), Global Moran’s I, Local Indicators for
Spatial Association (LISA) and K-Means (K-Means) clustering are used to determine
spatial differentiation and clustering trends. In addition, these analyses are supported
by demographic and socio-economic data, presenting the multidimensional structure
of urban inequalities.
LITERATURE REVIEW
https://www.periodicos.unimontes.br/index.php/rds/
127
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
From a sociopsychological perspective, as a result of interactions between
competition and cooperation, individuals often group around common identity values
such as ethnic and racial origin, religious belief, and socioeconomic status. Social
Identity Theory (TAJFEL; TURNER, 1979) and Self-Categorization Theory (TURNER et al.,
1987) offer frameworks to understand the formation and dynamics of these social
groups. The theories, propose that individuals derive their sense of self from their
membership in specific social groups. In return, groups determine the behavior of
individuals, their interactions with each other and their social positions.
According to Social Identity Theory, group membership provides individuals
with a sense of belonging and identity. This situation affects both intra-group and
inter-group relations. Identity formation, a dynamic process, leads individuals to
constantly compare their own groups with other groups. This comparison shapes
individuals’ attitudes towards competition and cooperation in line with the group
advantages and disadvantages they perceive. In addition, Self-Categorization Theory
elaborates on this process by emphasizing the cognitive mechanisms. According to
these mechanisms, individuals place others as well as themselves into categories. This
categorization influences their behavior in ways that support the interests of their in-
group. This reinforces divisions between groups because individuals tend to serve their
collective identities. This can lead to social consequences such as prejudice,
discrimination, and, in the extreme, dehumanization (TAJFEL AND TURNER, 1979;
HASLAM, 2006). The process of dehumanization occurs when other groups’ members
are viewed as morally or cognitively deficient, inferior, or threatening. Moreover,
outgroup members are viewed as inhumane and are therefore unable to demand the
resources they deserve. This can also lead to conflict (HASLAM, 2006).
From a sociological perspective, Henri Lefebvre's findings on the social
production of space also touch on these features of the urban segregation
phenomenon. These findings include the distinction between perceived space (physical
environment), designed space (space planned by authorities) and lived space (space of
daily experiences and resistance). In this context, space is not neutral. There is a
continuous production through social interactions, power struggles and ideological
practices. Class and social inequalities are reinforced in the production process, and
space hosts resistance and cooperation practices. In this context, space is a dynamic
https://www.periodicos.unimontes.br/index.php/rds/
128
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
structure with both product and producer characteristics (LEFEBVRE, 1991). According
to this framework, while space prepares the ground for the struggle to gain more
advantage among social groups, it also offers the opportunity to develop strategies to
combat these inequalities through solidarity.
Based on this, Edward Soja defines the third space, where physical, mental and
social spaces are interconnected. According to the concept, segregation is a product of
imagined and lived experiences beyond material reality. The concept of spatial justice
also put forward by Soja addresses the unequal distribution of resources and
opportunities. His work is important in understanding how racial identities are
constructed through space. He suggests that marginalized communities can achieve
gains in urban space through collective actions and alternative imaginations. In this
context, these communities can develop new spatial and social imaginations against
inequalities through collective actions and solidarity practices (SOJA, 2000). In other
words, the potential of cooperation to balance social competition is emphasized.
Pierre Bourdieu’s concepts of habitus and capital also focus on the economic,
social and cultural roots of the issue. Habitus (BOURDIEU, 1990) refers to the ways
individuals think, feel and act, shaped by their social positions and past experiences.
The concept of cultural capital (BOURDIEU, 1986) explains the critical role that
education, language, art and cultural practices play in maintaining and reproducing
differences between classes. This form of capital, which allows individuals to gain
advantage, also deepens spatial segregation. Because cultural capital improves the
mobilization of individuals within social hierarchies and provides access to more spatial
opportunities. Thus, space emerges as an area where social segregation and struggles
become concrete.
Loic Wacquant’s concepts of urban marginality and spatial stigma, which focus
on the experiences of marginalized communities, complement Bourdieu’s ideas.
According to these concepts, spatial segregation is a mechanism of social control and
exclusion beyond economic inequalities. Marginalized neighborhoods are stigmatized
with negative connotations that reinforce isolation. In this way, cycles of poverty are
perpetuated. Thus, by emphasizing the symbolic and material dimensions of
segregation, it is seen how racial identities are intertwined with spatial stigmatization.
As a result, stigmatized spaces are tools that serve to push marginalized communities
https://www.periodicos.unimontes.br/index.php/rds/
129
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
into disadvantaged positions. At the same time, stigmatized spaces push these
communities to develop solidarity practices (WACQUANT, 2008). Thus, the interaction
of social competition and cooperation dynamics in the spatial context is revealed
again.
Among the contributions specific to the southern sphere, Aníbal Quijano's
concept of coloniality provides a better understanding of the Latin American context.
According to concept, modernity is a historical context in which race is a fundamental
element of hierarchy, beyond the process of progress and development that is a claim
of the northern sphere’s perspective. The economic, cultural and social structures that
are the legacy of the colonial period also form the basis of today’s social inequalities.
This legacy has also left deep effects on spatial segregation processes. Thus,
colonialism is a form of domination that is shaped on the axis of race, identity and
culture, beyond economic exploitation. The structure it causes is constantly
reproduced through today’s cities (QUIJANO, 2000). Silvia Rivera Cusicanqui’s
decolonial perspective, in addition, addresses the issue at the center of indigenous
knowledge and resistance. Cusicanqui defines urban spaces as areas of conflict with
the concept of ch’ixi, which expresses the coexistence of opposites. The concept
emphasizes the resistance that indigenous and marginalized communities demonstrate
against dominant spatial logic in order to preserve their cultural identities in the
context of segregation (CUSICANQUI, 2012). Thus, Quijano and Cusicanqui’s decolonial
theories address the historical and epistemic dimensions of social competition.
Colonialism is an expression of the social competition between individuals and groups
throughout history, as well as an expression of alternative cooperation possibilities
against dominant knowledge and spatial systems.
In parallel, Walter Mignolo and Boaventura de Sousa Santos also emphasize
the importance of knowledge production and systems in social dynamics (MIGNOLO,
2011; SANTOS, 2014). Mignolo's concept of epistemic segregation explains the global
dominance of colonial knowledge systems and how they marginalize local alternative
forms of knowledge production. Western-centered epistemologies have shaped social
order as well as norms of knowledge production. Forms of knowledge outside the
West have been marginalized by being excluded, devalued, and labeled as false. In this
way, social structures, value systems, and power dynamics have also been
https://www.periodicos.unimontes.br/index.php/rds/
130
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
transformed. This separation in knowledge production also ensures the continuity of
existing inequalities. Santos supports this discussion with the concept of invisible
knowledge. Western-centered epistemologies systematically render other forms of
knowledge invisible. Thus, the questioning of the value of these forms of knowledge is
prevented. According to both philosophers, local and alternative knowledge systems
are a powerful tool against the system that produces existing inequalities. At this
point, the decisive effect of the control of information on inter-group competition and
cooperation configurations is seen, and the role of information production is further
underlined.
Frantz Fanon’s studies on colonialism and identity further deepen the analysis
of socio-spatial segregation. Fanon also draws attention to the identity problem of
individuals and groups affected by colonialism. Colonial systems transform their
identities by imposing their own norms and values on local individuals. In this process,
individuals experience internal conflicts in addition to external interventions. In this
context, it is emphasized that spatial segregation is also related to social perceptions,
identities and the subjective experiences of individuals. The racial and cultural
hierarchies that shape the identities of individuals and colonial discourses reproduce
the minds of individuals and the network of social relations. As a result, when
analyzing the phenomenon of segregation, psychological elements such as identity,
belonging and social harmony gain importance in addition to economic and political
dimensions (FANON, 1961). As a result, while the identities of individuals are shaped
by spatial segregation, competition and cooperation are experienced at individual and
social levels. While the legacy of colonialism leaves individuals at a disadvantage in
social competition, the role of cooperation is once again highlighted.
Lipsitz draws attention to another perceptual aspect of the issue. According to
him, racial discrimination and economic inequalities in cities are not only results but
also active elements of spatial organizations. While the white spatial imagination
idealizes the creation of safe, orderly and high-value areas, the idealization process
brings about the exclusion of other groups. Exclusion is accompanied by physical
boundaries as well as the direction of urban planning, housing policies and economic
investments. Thus, socioeconomic and cultural polarization is deepened. Reactions to
these processes bring us back to the concept of spatial justice. In order to ensure
https://www.periodicos.unimontes.br/index.php/rds/
131
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
justice, in addition to the reorganization of physical space, the systems that reproduce
social inequalities must be addressed and transformed in their historical contexts
(LIPSITZ, 2006).
David Harvey’s understanding of the social production of space is also parallel
to the theories of the pioneer philosophers (HARVEY, 2009). According to him, space is
not merely a physical area but also an area where social relations are shaped and
reproduced. The capitalist mode of production uses space as a tool for this purpose. In
addition to economic interests, social relations, power structures and ideological
practices are also effective in the production. While class and cultural positions among
individuals are reflected in space, these positions are also reproduced. Thus, space is
used as a tool that deepens class differences and reproduces inequality.
Saskia Sassen's theories on global cities and competition between cities are
also important contributions. Also according to her, with the driving force of
globalization, settlements that have become financial and trade centers have led to
deepening social inequalities and strengthening spatial segregation. Global cities have
become not only centers of economic activities but also centers of global governance,
communication and culture. As a result, the concentration of international capital and
labor has led to further polarization of urban space. The needs of global capital have
made spatial structures in cities and even certain regions economically valuable while
marginalizing other regions (SASSEN, 2001). As a result, it is seen that urban
segregation does not only stem from differences between local social classes but is
also driven by global economic structures. Thus, the reshaping of the social dynamics
mentioned within neo-liberal and global economic structures is emerging. As social
competition carried to a global scale intensifies, cooperation is also being reshaped.
When the above-mentioned ideas are synthesized, urban space can be defined
as a field of competition and cooperation where social groups compete for resources,
recognition and power. Social identities, whether racial, ethnic, class-based or cultural,
are not static, but dynamic phenomena shaped by historical processes and interactions
with space. Furthermore, segregation reflects and strengthens social power structures,
and on the other hand, gives rise to cooperation strategies. Therefore, urban
segregation is the reproduction of social inequalities related to the identities and
struggles of competing groups as well as physical structures. At the end of the process,
https://www.periodicos.unimontes.br/index.php/rds/
132
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
groups that are placed at the upper levels of the social hierarchy in line with their
strategies become advantageous, while those whose mobilization is restricted and
imprisoned at the lower levels become disadvantaged. The tension between these two
dynamics that shape social transformation and resistance plays a decisive role in the
evolution of social structures.
These dynamics are intertwined with structural racism, income concentration
mechanisms, laws, urban planning, urban transformation projects, and real estate
speculation. Urbanization processes create mechanisms that reinforce structural
racism, while discriminatory housing policies confine certain racial or ethnic groups to
areas with limited access to infrastructure, educational and economic opportunities. In
addition, these areas are socially and economically excluded through spatial
stigmatization. Moreover, hierarchies related to historical context reproduce
themselves in space. As a result, social inequalities are further deepened (WACQUANT,
2008; HARVEY, 2009).
Legal systems also play an important role in perpetuating inequalities. They
tend to protect the interests of advantaged groups at the top of the social hierarchy.
For example, expropriation laws often target disadvantaged groups located in poor
neighborhoods, prioritizing capital accumulation over social equality. Legal frameworks
inherited from colonialism further deepen inequalities by excluding local knowledge
systems and community-based approaches (QUIJANO, 2000; MIGNOLO, 2011;
SANTOS, 2014).
Urbanization under the influence of globalization also increases social injustices
by concentrating wealth in certain regions. While city centers generally become areas
where high-income advantaged groups live and economic opportunities are
concentrated, low-income disadvantaged groups are pushed to the periphery. In the
process, urban planning becomes a tool that shapes social inequalities spatially. This
tool generally prioritizes the interests of advantaged groups. Infrastructure
investments remain limited in areas where disadvantaged groups are settled. The
needs and information systems of these communities are not taken into consideration
as much as they should be in planning processes (SASSEN, 2001; HARVEY, 2009). As a
result, spatial injustices become permanent and social hierarchies are reproduced in
urban space (LEFEBVRE, 1991).
https://www.periodicos.unimontes.br/index.php/rds/
133
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Moreover, urban transformation projects, which are tools that accelerate
capital accumulation, further victimize disadvantaged groups by prioritizing economic
gain over social equality. These groups are pushed more violently to the periphery of
the city, while the positions of advantaged groups are strengthened. Thus, the
identities and cultural assets of disadvantaged local communities are destroyed
through both spatial and cultural transformations (SOJA, 2000). In the process, real
estate speculation also increases inequalities by commodifying urban space. High
housing prices caused by speculation make it even more impossible for low-income
groups to live in the center. While speculative investments consume resources that
could be allocated to public infrastructure and affordable housing projects,
international capital movements further segregate city centers and ignore the needs of
local people (SASSEN, 2001; HARVEY, 2009).
Map 1: São Paulo State and its location in Brazil.
Obs. The port of Santos has maintained its importance since the early development of the settlement. It
has undertaken important roles during different periods, from the transportation of goods abroad to the
arrival of immigrants to the state. Source: RIBEIRO (1924)
https://www.periodicos.unimontes.br/index.php/rds/
134
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
In Metropolitan São Paulo, which is apt to reflect this complexity of the issue,
the diversity dates back to 1554. The city of São Paulo was first established as a
mission by Portuguese Jesuit priests (FREYRE, 1987). In Brazil, where economic
activities were concentrated in coastal areas such as Pernambuco and Bahia, São Paulo
exhibited a rural character at that time (MARTINS, 2000). In the 17th century,
explorers called Bandeirantes, who were in search of gold and precious mines, used
the city of São Paulo as a station while organizing expeditions to the interior. During
these expeditions, the indigenous people were either enslaved to work or killed
(PRADO, 1994). In the 17th and 18th centuries, the need for labor began to be met by
slaves brought from Africa (BETHELL, 1989). With the decline of the indigenous
population and the institutionalization of slavery in the Portuguese colonies, the
economic structure of São Paulo began to change. In the 18th century, the fact that
São Paulo was a transit point due to the gold mines in the state of Minas Gerais
reinforced the importance of the settlement for Portuguese colonialism (BOXER,
1962).
Since the 19th century, Brazilian cities, including São Paulo, have become
centers of production, migration and consumption (LANNA, 1996). The economic
structure has been based on agricultural production shaped by European capitalist
economies, where rural oligarchies settled in São Paulo and Minas Gerais holding the
power. During this process, the agricultural production has also been sold abroad
(FAUSTO, 2000). Additionally, São Paulo's railway network contributed to
industrialization and the development of some urban settlements that became
production centers under the control of the European market (see map 1). The existing
socio-economic structure has also caused urbanization to intensify (CANO, 2012).
After the abolition of slavery in 1888, the coffee industry’s need for a large
workforce was met by large waves of migration from Europe. Italians, Portuguese,
Germans and Spanish flocked to São Paulo (HOLLOWAY, 1980). Thus, during the
expansion of the settlement these communities established their own neighborhoods
(LEVINE, 1999). In addition, Japanese immigrants, who found a place for themselves
especially in the agricultural sector, were another group that contributed to the social
fabric of the city (TSUNECHIRO AND PINO, 2008). In the mid-20th century, economic
difficulties, drought and social inequalities in the Northeastern states of Brazil caused
https://www.periodicos.unimontes.br/index.php/rds/
135
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
large waves of internal migration to São Paulo. Immigrants, who were generally of
Afro-Brazilian origin, increased the group diversity. In this period also historically
marginalized indigenous, black and brown populations were integrated into the social
structure (AMARAL, 2013).
Today, the state of São Paulo, one of the 27 federative units that make up
Brazil, offers a dynamic structure with a population of 44,411,238, 645 municipalities
and a total gross income of R$ 343,634,435,413.40. According to 2022 data, 57.78% of
those living in the state of São Paulo identified themselves as white, 32.96% as brown,
7.99% as black, 1.16% as yellow and 0.11% as indigenous. Metropolitan São Paulo,
which includes São Paulo as capital and 38 separate municipalities, also hosts a
population of 20,743,587. The largest municipalities in this dynamic structure are São
Paulo with a population of 11,451,245 and Guarulhos with a population of 1,291,784.
There are also municipalities with high populations such as São Bernardo do Campo
(810,729), Santo André (723,889) and Mauá (472,912) in the ABC Region, where
industrial production is concentrated (IBGE, 2024).
Thus, Metropolitan São Paulo has been the subject of several studies
addressing the issue. These studies reveal that the black, brown and indigenous
populations are subject to a systematic process of exclusion due to historical and
structural dynamics. Thus, these groups face social, economic and spatial inequalities
within the city. Historically, immigrants from Europe and Asia have had more economic
and social opportunities, while others have been pushed into low-wage jobs, the
informal sector and the periphery. Black and brown populations in particular are
concentrated in informal settlements, social housing projects and low-income
neighborhoods. Indigenous communities are often exposed to environmental
injustices and struggle to preserve their identities (MARQUES and RODRIGUES, 2013;
BARROS, MEDEIROS and MORAIS, 2016; FRANÇA, 2020; PRETECEILLE AND CARDOSO,
2020).
Marginalization is reinforced by urban policies and structural racism. According
to current police violence statistics, black and brown young men are the most
targeted. Racial inequalities exist in the criminal justice system (SOUDAPAZ, 2024). In
addition, the social mobility of these groups is limited. Inequalities in access to quality
education and health services perpetuate the cycle of poverty. Gentrification
https://www.periodicos.unimontes.br/index.php/rds/
136
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
processes also increase marginalization. Urban transformation projects are often
implemented in a way that further excludes low-income black and brown
communities. Thus, spatial segregation deepens. As a result, black, brown and
indigenous populations are excluded both spatially and economically and politically
(CALDEIRA, 1997; PEARLMAN, 2010; VILLAÇA, 2011).
METHOD
The study uses an approach that integrates statistical and spatial analyses to
reveal socio-economic and racial patterns. The methodology consists of data
preparation, summary statistics, social concentration, spatial auto-correlation, and
socio-economic cluster analysis steps. The municipality-based dataset used in the
analysis includes demographic and socio-economic variables (IBGE, 2022; OBSERVA
SAUDE, 2022; SEADE, 2022). This set and the geographic information system files
showing municipal boundaries were obtained from the IBGE (2022), the official
statistics agency of Brazil. Finally, analyses and visualizations are performed using
python libraries and packages.
In addition to summary statistics (WEISBERG, 2005), Location Quotient (ISARD,
1956), Global Moran’s I (MORAN, 1950), Local Indicators of Spatial Association
(ANSELIN, 1995) and K-Means Clustering (MACQUEEN, 1967) methods were used for
detailed analysis. Each method provided outputs from different perspectives to
understand the relationships between the racial groups comprehensively.
The population structure of each municipality is represented by the official
classification categories of indigenous, black, brown, yellow and white populations.
Summary statistics provide a summary of the distribution, central tendency, and
variability of data. In the context of urban segregation, they can be used to understand
the general profile of the population before complex analyses. For example, they can
provide a quick look at income inequalities, the concentration of racial groups, or the
distribution of income levels in certain areas (WEISBERG, 2005). In the study, summary
statistics were also used to summarize the main features of the data set and provide
an overview. The data set contains geographic size, population, population density,
gender ratio, birth and death rates, infant deaths (under 1 year), married population,
https://www.periodicos.unimontes.br/index.php/rds/
137
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
household numbers, child dependency ratio, elderly dependency ratio, GDP, formal
employment numbers, education levels, scientific, technical, educational, cultural,
recreational and health organizations, international organizations, hospitalization
numbers, health facility numbers, health expenditure per capita, the number of
households with water supply, and with connection to sewage system.
Location Quotient (LQ), a non-spatial measure, is a ratio that measures how
densely populated social groups are in a particular area within a region compared to
the entire region. It is generally used to examine the distribution of socio-economic or
demographic groups (LI and GOU, 2020). In the study, it also helped to understand the
spatial clustering densities of the groups considered. The coefficients obtained for each
group revealed whether a particular racial group was over-represented (LQ > 1) or
under-represented (LQ < 1) in the general population distribution.
Moran’s I statistic is used to measure spatial auto-correlation. The method can
be used to test the spatial dependence of a particular socio-economic variable, such as
income level or racial distribution, across geographic areas. In urban segregation
studies, Moran’s I is used to understand whether similar socio-economic groups are
clustered in certain areas (TORRES and BICHIR, 2009; CUNHA and JIMENEZ, 2009;
FLORES and WILSON, 2009; GROISMAN and SUAREZ, 2009). A positive and significant
Moran’s I value indicates a strong spatial relationship between similar values. This
indicates that certain social groups are spatially separated. The existence of groups
formed around common identity elements clustered in certain areas is analyzed with
Moran’s I. The main results also included expected I values, and global and local p-
values. The expected I value is the value expected to be obtained from a random
distribution of Moran’s I. Deviations from this value helps understanding the spatial
structure. In addition, the global p-value is used to test whether a cluster is statistically
significant. A low value indicates that the clustering is not random. Finally, the local p-
value shows whether there is significant clustering in certain regions.
Local Indicators of Spatial Association (LISA), a local version of Moran’s I, helps
determine the degree and spatial patterns of segregation between different areas in
urban segregation analyses. Although being a different analysis, it uses a local iteration
of Moran’s I with weighting in cluster determination (CUNHA ET AL., 2009; FLORES,
2009; NIELSEN and HENNERDAL, 2017; POULSEN, JOHNSTON, and FORREST, 2010;
https://www.periodicos.unimontes.br/index.php/rds/
138
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
GIBBONS et al., 2020). In the study, it was used to determine spatial heterogeneity.
Queen contiguity weights were created to define neighboring municipalities for spatial
autocorrelation analysis. As a result, spatial clusters in areas where certain racial
groups are concentrated were determined. A more comprehensive understanding was
aimed by using Moran’s I and LISA, which complement each other.
K-Means Clustering, on the other hand, defines common clusters by grouping
data points according to their similarities through unsupervised learning. The method
is often used in large data sets to separate data into meaningful subgroups (DONEGAN
and TAVARES, 2024; KILANI and DAHER, 2024). After determining the optimum
number of clusters (k), each data is assigned to the closest cluster and the mean points
(centroids) of the clusters are continuously updated. K-Means was used to determine
areal clusters according to socio-economic and demographic characteristics. The
variables include gross domestic product (GDP), formal employment, nominal average
salary, total dependency ratio and racial group percentages. Before clustering, the data
was standardized using the Standard Scaler to reduce the negative effects of different
units in which the variables were expressed. Thus, all variables used were on the same
scale. Elbow Method (THORNDIKE, 1953) and Silhouette Scores (ROUSSEEUW, 1987)
were used to determine the optimum number of clusters. Then, K-Means clustering
was performed and municipalities with similar socio-economic and racial profiles were
grouped into four separate clusters. Thus, it is aimed to better understand the reasons
for the segregation.
RESULTS
As a result of the summary statistics, a heterogeneous spatial, demographic
and socio-economic structure was detected in the study area. The inequalities are
reflected in the uneven distribution of population, land, economic output, access to
health services and basic infrastructure. Below is the municipality-based general
profile of the Metropolitan São Paulo derived from mentioned demographic and socio-
economic indicators.
The average geographic size of the municipalities is 203.77 km², while the
standard deviation value is 268.64 km². This shows a remarkable size diversity. The
https://www.periodicos.unimontes.br/index.php/rds/
139
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
population distribution is even more unbalanced. The average population is 531,558.
However, a very high standard deviation (1,814,200.14) indicates extreme differences
in population sizes among municipalities. Additionally, the average population density
is 3,480.42 people/km². The high standard deviation value (3,941.86) here also reflects
the density changes between urban, rural or semi-rural areas. This situation can be
thought to be due to factors such as urbanization history, economic activities and
infrastructure development. In contrast, the gender ratio is relatively consistent. There
are an average of 94.27 males for every 100 females. The standard deviation (3.40)
indicates only a slight unbalanced distribution. Moreover, birth and death rates are
relatively balanced throughout the region. This relative balance indicates a moderate
population growth. Although there is an average of 71 infant deaths per 1,000 live
births, the distribution of infant mortality reveals inequalities with a standard deviation
of 224.13, especially in access to health services. This suggests deeper socio-economic
and public health differences.
The distribution of married population is also unbalanced, in line with other
demographic data. While most municipalities have relatively fewer married couples, a
small group of them are more densely populated with them (standard deviation:
8949.60). Household structures also show significant differences. A small number of
municipalities have the majority of households. The average number of households is
224,978. Furthermore, the standard deviation (791,463) highlights the large
differences. While the dependency ratios for children and the elderly show some
balance, they also indicate differences in demographic pressures on the working-age
population. Some municipalities have much higher proportions of elderly people than
others. Minimum and maximum values for the elderly dependency are 26.26 and 3.27,
respectively.
The average municipal GDP is approximately R$35.6 million. However, the large
standard deviation (R$132M) indicates significant economic inequalities. Only a few
municipalities contribute disproportionately to the overall economy. In addition,
formal employment numbers follows a similar pattern. On average, there are 192,363
formal jobs in each municipality. However, the very high standard deviation
(870,176.4) highlights the unequal distribution of employment opportunities.
https://www.periodicos.unimontes.br/index.php/rds/
140
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Education levels (min. 0.41 and max. 0.72) vary. In addition, scientific,
technical, educational, cultural, recreational and health organizations are concentrated
in certain municipalities. There are significant differences in the total number of
companies with a standard deviation of 101,478.90. Moreover, international
organizations are concentrated only in the city of São Paulo. Thus, standard deviation
values highlight the uneven development of civil infrastructure. Health care use and
expenditures are also uneven. Hospitalization rates and access to health facilities are
concentrated in a few municipalities. The average number of patients is 26,188.
Additionally, the high standard deviation for this variable (87,136.65) underlines the
inequalities in access to health services and infrastructure. Although per capita health
expenditure varies at a moderate level (min. 446.84 and max. 4,281.96), it is still
biased towards some municipalities. Finally, standard deviation values for access to
water supply systems also indicate significant differences in basic infrastructure. The
standard deviation value for the households without connection to network and with
connection to sewage system are 4,273.22 and 651,040.57, respectively. The existence
of many municipalities with poor access deepens public health and environmental
inequalities.
LQ results (see table 1) for racial groups provide clues to the interplay between
socio-economic factors, cultural dynamics and historical contexts. Economic pressures,
cultural preferences and historical migration patterns appear to have contributed to
the segregation and integration of all these groups. Indigenous populations are more
concentrated in peripheral and semi-rural municipalities such as Pirapora do Bom
Jesus, Guararema and Itapecerica da Serra. These areas should be offering
opportunities for shelter, proximity to natural resources and cultural preservation. In
contrast, these populations are less concentrated in urbanized municipalities such as
Salesópolis, Itapevi and Osasco. These ones might be less accessible or desirable for
indigenous residents due to cultural factors and economic pressures. Due to historical
processes, cultural or social networks that would support indigenous populations may
not have developed. São Paulo and Guarulhos, where the population is moderately
represented, indicate a relative integration. Thus, it can be seen that despite the
presence of indigenous communities in the metropolitan area, their integration and
segregation vary significantly.
https://www.periodicos.unimontes.br/index.php/rds/
141
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
The results for the black population presents a complex demographic structure.
This population is concentrated in Embu das Artes, Francisco Morato and Diadema.
These municipalities must have strong community networks serving the group, thus
standing out as concentration areas with cultural resources that increase the group’s
resilience. On the other hand, Salesópolis and São Caetano do Sul have lower density
values, indicating important obstacles, likely systemic inequalities.
Table 1: LQ coefficients for racial groups in the municipalities.
Source: Prepared by the author.
Embu das Artes, Francisco Morato and Pirapora do Bom Jesus are important
centers for the brown population. Apparently, these areas can provide the group with
housing, cultural vitality and community cohesion. On the other hand, the low
densities encountered in São Caetano do Sul and Salesópolis suggesting socio-
economic barriers that prevent the settlement and growth of this group. These
Municipality
Aruja 0,978 0,86 1,083 1,01 0,977
Barueri 0,524 0,85 1,09 0,431 0,993
Birit iba Mirim 0,669 0,629 0,937 2,264 1,081
Caieiras 1,093 0,987 1,024 0,171 1,019
Cajamar 0,61 0,899 1,146 0,207 0,953
Carapicuiba 0,829 1,147 1,2 0,183 0,87
Cot ia 0,879 0,935 1,101 0,684 0,957
Diadema 0,559 1,112 1,21 0,338 0,862
Embu das Artes 0,743 1,386 1,347 0,237 0,723
Embu Guacu 1,06 0,898 1,211 0,426 0,892
Ferraz de Vasconcelos 0,946 1,225 1,255 0,137 0,811
Francisco Morato 0,647 1,358 1,377 0,071 0,715
Franco da Rocha 0,548 1,101 1,174 0,106 0,839
Guararema 1,32 0,655 0,94 0,755 1,115
Guarulhos 0,929 0,977 1,14 0,582 0,919
Itapecerica da Serra 1,214 1,216 1,277 0,338 0,793
Itapevi 0,329 1,101 1,41 0,113 0,738
Itaquaquecetuba 1,019 1,104 1,38 0,169 0,756
Jandira 0,537 1,066 1,201 0,261 0,883
Juquit iba 0,947 0,626 1,063 0,316 1,048
Mairipora 0,805 0,682 1,021 0,464 1,063
Maua 0,679 0,933 1,125 0,244 0,955
Mogi das Cruzes 0,903 0,874 0,921 2,3 1,042
Osasco 0,561 0,98 1,057 0,502 0,982
Pirapora do Bom Jesus 1,413 1,021 1,358 0,167 0,782
Poa 0,581 1,21 1,081 0,315 0,934
Ribeirao Pires 1,282 0,741 0,979 0,619 1,073
Rio Grande da Serra 0,795 1,105 1,291 0,241 0,811
Salesopolis 0,201 0,255 0,607 0,441 1,411
Santa Isabel 0,517 0,604 0,891 0,58 1,163
Santana de Parnaiba 0,822 0,75 1,033 0,604 1,043
Santo Andre 0,633 0,661 0,757 0,813 1,232
Sao Bernardo do Campo 0,979 0,746 0,906 0,903 1,119
Sao Caetano do Sul 0,853 0,374 0,416 1,255 1,501
Sao Lourenço da Serra 0,523 0,629 1,063 0,808 1,032
Sao Paulo 1,171 1,039 0,934 1,302 1,027
Suzano 0,725 1,004 1,157 1,389 0,887
Taboao da Serra 0,938 1,284 1,126 0,557 0,883
Vargem Grande Paulista 0,515 0,725 0,957 1,302 1,077
Indian
population Black
population Brown
population Yellow
population White
population
https://www.periodicos.unimontes.br/index.php/rds/
142
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
findings suggest the specific difficulties faced by these communities in the areas of
housing, employment and cultural support.
There are also clear spatial variations in the yellow population distributions.
The over-representation in Biritiba Mirim and Mogi das Cruzes suggests the existence
of racial pockets or historical settlements that have persisted over time. Probably,
these municipalities offer cultural, social or economic environments that is attractive
to this group due to social networks. However, municipalities with very low LQ values
such as Ferraz de Vasconcelos and Francisco Morato show a limited presence. This
might be related with the socio-economic factors that not matching the needs of this
group. Finally, the relatively balanced distribution in São Bernardo do Campo and
Osasco suggests integration.
Finally, the high LQ values of the white population in São Caetano do Sul, Santo
André and Salesópolis suggest that historical migration patterns, socio-economic
factors and housing availability contribute to concentration. It is noteworthy that these
municipalities generally offer higher living standards and more developed
infrastructures. In contrast, municipalities such as Francisco Morato and Embu das
Artes have lower densities. These areas are more populated by the other groups.
Peripheral areas and low-income areas tend to host relatively fewer members of the
white population than other areas.
The LISA analysis for the percentage of indigenous population (see map 2)
shows a negative Global Moran’s I value (−0.095), indicating slightly negative spatial
auto-correlation. This suggest that municipalities with higher densities are more
dispersed. The global p-value is 0.259, indicating that the distribution is not statistically
significant. Local p-values are above 0.05 in many municipalities. Thus, there is no
significant clustering. However, some municipalities show lower values, pointing out
isolated areas of concentration. Overall, these findings shows that the indigenous
population is dispersed within the larger community of the metropolitan area.
A value of 0.312 for the black population presents a moderate positive spatial
auto-correlation. This suggests that municipalities with higher densities tend to cluster
together (see map 3). The expected I value is −0.026 and the global p-value is 0.001.
These confirm the statistically significant clustering. Local p-values also highlight this
significancy in some municipalities with values around 0.004 and 0.015. However, the
https://www.periodicos.unimontes.br/index.php/rds/
143
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
majority of the municipalities do not have significant p-values, indicating a more
diverse demographic distribution. As a result, the municipalities where this group is
concentrated form a certain spatial structure.
Map 2: LISA clusters for Indigenous population.
Source: Prepared by the author.
Map 3: LISA clusters for black population.
Source: Prepared by the author.
https://www.periodicos.unimontes.br/index.php/rds/
144
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
With a value of 0.241, the brown populations show moderate positive spatial
auto-correlation (see map 4). This suggests that municipalities with higher percentages
are generally clustered together. There are significant spatial patterns. The statistical
significance of these patterns is supported by the global p-value of 0.008. Local p-
values of 0.015 and 0.023 also support this significancy in several municipalities.
Conversely, many municipalities exhibit local p-values above 0.05, suggesting more
integrated demographic distributions. As a result, a significant concentration is
revealed also for this group.
Map 4: LISA clusters for brown population.
Source: Prepared by the author.
The yellow population also exhibits positive spatial auto-correlation with a
Global Moran’s I value of 0.180 (see map 5). Municipalities that host higher yellow
populations tend to cluster together with a certain pattern. Additionally, the global p-
value is 0.029. The clustering is statistically significant. Many municipalities with local
p-values below 0.05 support the result. In particular, areas with local p-values
indicating strong clustering reflect socioeconomic or cultural factors affecting the
distribution of the group.
https://www.periodicos.unimontes.br/index.php/rds/
145
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Map 5: LISA clusters for yellow population.
Source: Prepared by the author.
Map 6: LISA clusters for white population.
Source: Prepared by the author.
https://www.periodicos.unimontes.br/index.php/rds/
146
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Finally, a positive value (0.251) for the white group indicates that this
population tends to be more concentrated than dispersed (see map 6). The global p-
value of 0.01 also confirms that the clustering is significant. Further, local p-values
highlights this significancy, particularly in some municipalities below 0.05. This
clustering may reflect the influence of socioeconomic factors, historical settlement
patterns, or policies that facilitate residential segregation.
K-Means cluster analysis also reveals distinct patterns in terms of various
socioeconomic and demographic indicators (see table 3). The optimum number of
clusters is defined as four. These clusters, characterized by their centroids, provide
more information about the socioeconomic conditions and racial compositions of the
areas.
Table 3. Clustered Municipalities
Source: Prepared by the author.
Cluster 0 exhibits moderate values for the most socio-economic indicators.
Gross domestic product (GDP) and formal employment are slightly below average,
while nominal average wages are significantly higher than other characteristics. This
suggests potential wealth inequality among municipalities in the cluster. According to
racial composition data, the indigenous population (−0.444) is significantly
underrepresented, while the brown (0.361) and black population (0.433) are
moderately represented. Thus, this cluster includes municipalities with a complex
demographic structure and moderate economic activity. Although it constitutes an
Cluster 0 1 2 3
Municipality Barueri Biritiba Mirim Aruja Sao Paulo
Cajamar Mogi das Cruzes Caieiras
Cotia Salesopolis Carapicuiba
Diadema Santa Isabel Embu Guacu
Embu das Artes Santo Andre Ferraz de Vasconcelos
Franco da Rocha Sao Caetano do Sul Francisco Morato
Guarulhos Sao Lourenço da Serra Guararema
Itapevi Vargem Grande Paulista Itapecerica da Serra
Jandira Itaquaquecetuba
Maua Juquitiba
Osasco Mairipora
Sao Bernardo do Campo Pirapora do Bom Jesus
Suzano Poa
Taboao da Serra Ribeirao Pires
Rio Grande da Serra
Santana de Parnaiba
https://www.periodicos.unimontes.br/index.php/rds/
147
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
economically active community, there are also challenges in income inequality and
racial representation.
Cluster 1 consists of municipalities that reflect low GDP, formal employment,
and nominal average salary. All indicators are well below average (−0.202, −0.171, and
−0.780, respectively). The racial demographics here are particularly underrepresented,
with brown (−1.308) and black populations (−1.269). This composition also points to
systemic inequality and social mobility. The residents in the cluster, likely struggle with
economic hardship and lack of resources, offering limited opportunities for community
advancement.
Cluster 2 has low GDP and employment figures, as well as moderate salary
levels (−0.240, −0.215 and −0.371 respectively). However, it has significantly higher
representation of the indigenous population (0.672). Also, brown (0.383) and black
population (0.225) are moderately represented. The higher presence of the indigenous
group suggests that the cluster may contain municipalities with specific cultural or
economic characteristics for this group. Finally, despite the large challenges in the
area, there are also economic opportunities.
Map 7: K-Means clusters of municipalities
Source: Prepared by the author.
https://www.periodicos.unimontes.br/index.php/rds/
148
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Cluster 3 contrasts with the others by displaying exceptionally high values for
GDP, formal employment and nominal average salary. In this cluster, the indigenous
population (1.358) has a relatively strong representation, while the black population
(0.481) has a moderate and the brown population (−0.739) has a lower representation.
This cluster includes the economic centers of the metropolitan area, which offer high
employment and income levels. The significant presence of the indigenous population
suggests the existence of potential cultural and economic networks that supports
them.
As a result, K-Means cluster analysis highlights the diversity of socioeconomic
conditions and racial composition within the Metropolitan Region (see map 7). While
some areas are economically dynamic, others struggle with inequality. Cluster 0
represents moderate socioeconomic conditions. It offers opportunities for policies to
reduce economic inequalities and promote racial inclusiveness. Cluster 1, where all
racial groups are significantly underrepresented, needs critical interventions to
alleviate economic hardship and improve access to resources. Intensive efforts are
needed to address systemic barriers. Cluster 2 presents mixed results. It offers growth
and development potential, particularly for the indigenous population. This cluster
could benefit from policies that promote cultural diversity while improving local
economic conditions. Finally, Cluster 3 represents the economic elite of the region. In
addition to strategies to sustain growth, approaches should be developed to share
economic prosperity across all groups.
CONCLUSION
In Metropolitan São Paulo heterogeneous demographic, social, economic and
spatial structures reveal that intense competition continues among racial groups.
While existing inequalities are both the result and the guarantor of relatively intense
access of certain groups to various services and opportunities, it is observed that
disadvantaged groups are eliminated at different levels.
This strengthens the of economic and infrastructural opportunity
concentration, especially in a small number of municipalities, while the inhabitants of
the remaining municipalities are condemned to struggle to access the resources. The
https://www.periodicos.unimontes.br/index.php/rds/
149
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
differences between urban, semi-rural and rural areas also play an important role in
maintaining the imbalances in access to infrastructure and services in the future.
Population density and demographic structures are also concentrated in
certain municipalities as a result of urbanization processes. In these areas, the pressure
on resources increases. In return, competition also intensifies. On the other hand, rural
or semi-rural areas with lower densities is left behind in terms of both economic
development and infrastructure. Thus, the population is concentrated around
economic opportunities, and the peripheral municipalities are developed to a limited
extent, while the central municipalities are developed disproportionately.
Inequalities in access to health services also demonstrate the impact of
competition between groups on public health. While access to these services is limited
in certain regions, there is disproportionate access in some other. This additionally
increases the competition and maintains the unbalanced distribution. Furthermore,
the concentration of relatively higher levels of education in certain municipalities also
leads to similar results. The dynamics between urbanization processes and economic
development also direct the access to these opportunities. Since educational
competition is a factor that strengthens socio-economic positions, it facilitates the
maintenance of advantaged groups’ positions.
The Location Quotient (LQ) analysis results also indicate distinct spatial
patterns. Municipalities with high LQ values for marginalized groups points out
cooperation where intra-group community networks provide cultural and social
support that serve to enhance resilience. However, the overall picture reflects deep
inequalities. Intra-group cooperation does not transcend the broader dynamics of
competition that shape access to resources, infrastructure, and opportunities across
the landscape. Thus, the current socio-spatial structure reflects the persistence of
historical competition between groups driven by systemic constraints.
Furthermore, the analysis of Local Spatial Association Indicators (LISA) reveals
varying degrees of spatial auto-correlation for different populations. There are
significant differences in the spatial patterns. Indigenous populations show limited
clustering, highlighting the strength of the socio-economic barriers and limited support
for cultural preservation. The remaining groups show varying degrees of positive
spatial auto-correlation. Black and white populations are particularly prominent among
https://www.periodicos.unimontes.br/index.php/rds/
150
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
them, reflecting cooperation within groups. Spatial consolidation of social networks
might be playing an important role within these groups. However, black and brown
populations are generally concentrated in marginal or peripheral areas, while the
white population is concentrated in more advantaged municipalities. This suggests that
the white population still benefits from historical and ongoing privileges. These spatial
patterns are further evidence of the persistent effects of socioeconomic inequality and
systematic inequalities that shape the urban landscape.
K-Means analysis also confirms previous results. Four clusters representing
diverse socioeconomic conditions and racial compositions provide information on the
socio-spatial organization correlated with economic indicators. The remarkable
differences among clusters call for policy interventions. The focus should be on
reducing deep-rooted inequality, promoting integration, and ensuring that economic
development benefits all communities.
The identified segregation, which also confirms previous studies, is not simply a
matter of physical distance between social groups. As Quijano (2000) argues, the past
colonial experiences continue to shape the region. The racial hierarchies established
during the colonial period continue to exist today, especially where black and
indigenous populations are confined to the peripheries. It is seen that the mobility of
non-white populations and their access to urban resources were restricted.
Cusicanqui’s (2010) concept of ch’ixi can be used to understand the tension
between the advantaged and dis advantaged groups. The area is a site of cultural
resistance where marginalized groups struggle to preserve their identities and reclaim
their rights. In this context, Mignolo’s (2007) concept of epistemic disjunction also
highlights how local knowledge systems, especially those of marginalized groups, are
often excluded. The neglect of these knowledge systems in urban planning processes
must have contributed to the perpetuation of spatial and social inequalities.
Urban marginality and spatial stigma (Wacquant, 2008) are also relevant here.
Disadvantaged groups are likely marginalized through processes and physical isolation.
This leads to a lack of access to quality services, housing, and infrastructure while
reinforcing the social stigma that marks these areas with negative connotations. Thus,
spatial segregation can be considered as a continuous process of exclusion.
https://www.periodicos.unimontes.br/index.php/rds/
151
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Metropolitan São Paulo is also an example of capitalist urbanization that leads
to the concentration of wealth in certain areas. As described by Harvey (2008), the
region should be considered as a mechanism that reinforces and reproduces social
inequalities. Advantaged groups reside in areas with access to economic opportunities
and services, while marginalized groups are pushed to the periphery. In return, this
spatial distribution should be expected to increase income inequality, as certain areas
are invested more while others remain underdeveloped. This also reinforces class-
based spatial divisions. It is very likely that existing legal and policy frameworks are
inadequate to meet the needs of marginalized communities. This issue must also be
addressed in the context of existing patterns of segregation.
Moreover, as argued by Sassen (2001), global capital has positioned the city of
São Paulo and the metropolitan region in a prominent position in the global urban
network. Thus, the region is likely influenced by international capital interests rather
than needs based on local interests. The flow of capital into real estate markets must
have contributed to increasing inequality. As noted by Harvey (2008) and Wacquant
(2008), the speculative nature of real estate development transforms urban space into
a commodity, prioritizing investment returns over the well-being of residents. This
process can be considered as an example of the exploitation of urban space as a way
of reproducing social hierarchies.
As a result, each of the above topics can be examined separately. However, this
study focuses the determining role of social dynamics in the Metropolitan São Paulo.
The results underline that the struggle of different social groups to access resources
increases imbalances throughout the area. This further reinforces existing inequalities.
It shows areas where social competition is intense with the uneven distribution of
resources such as economic, education and health services. Social cooperation and
socio-economic integration processes, on the other hand, are limitedly effective as a
by-product of this competition. A more balanced resource distribution policy should be
developed between urban, semi-rural and rural areas. Strengthening cooperation will
enable resistance to the intensity of competition and reduction of socio-spatial
injustices. In this way, all social groups living in both the center and the periphery
might have more equal access to urban resources.
https://www.periodicos.unimontes.br/index.php/rds/
152
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
REFERENCES
LANSELIN, Luc. “Local Indicators of Spatial Association—LISA”. Geographical
Analysis, v. 27, n. 2, p. 93-115, 1995.
BARROS, Ana Paula Borba Gonçalves; MEDEIROS, Valério Augusto Soares de;
MORAIS, Maria da Piedade. “A configuração espacial para o diagnóstico dos
assentamentos precários no Brasil.” In: MORAIS, Maria da Piedade; KRAUSE, Cleandro;
LIMA NETO, Vicente Correia (Eds.). Caracterização e tipologia de assentamentos
precários: estudos de caso brasileiros. Brasília: Instituto de Pesquisa Econômica
Aplicada – IPEA, 2016, p. 95-125.
BETHELL, Leslie. Colonial Brazil. Cambridge: Cambridge University Press, 1989.
BOURDIEU, Pierre. The Logic of Practice. Stanford: Stanford University Press, 1990.
BOURDIEU, Pierre. The Forms of Capital. In: RICHARDSON, John G. (Ed.). Handbook of
Theory and Research for the Sociology of Education. New York: Greenwood, 1986. p. 241–258.
BOXER, Charles R. The Golden Age of Brazil: Growing Pains of a Colonial Society,
1695-1750. Berkeley: University of California Press, 1962.
CANO, Wilson. Ensaios sobre a Formação Econômica Regional do Brasil.
Campinas: Unicamp, 2012.
CALDEIRA, Teresa Pires do Rio. “Enclaves Fortificados: A Nova Segregação
Urbana.” Novos Estudos, n. 47, p. 155-176, 1997.
CUNHA, José Marcos Pinto da; JIMENEZ, Maren Andrea; PEREZ, José Rafael
Ruiz; ANDRADE, Cibele Yahn de. “Social segregation and academic achievement in
state-run elementary schools in the municipality of Campinas, Brazil.” Geoforum, v. 40,
n. 5, p. 873–883, 2009.
CUNHA, José Marcos Pinto da; JIMÉNEZ, Maren Andrea. “The Process of
Cumulative Disadvantage: Concentration of Poverty and the Quality of Public
Education in the Metropolitan Region of Campinas.” In: ROBERTS, Bryan R.; WILSON,
Robert H. (Eds.). Urban Segregation and Governance in the Americas. New York:
Palgrave Macmillan, 2009, p. 170-185.
CUSICANQUI, Silvia Rivera. Ch’ixinakax utxiwa: A Reflection on the Practices and
Discourses of Decolonization. South Atlantic Quarterly, v. 111, n. 1, p. 95–109, 2012.
DONEGAN, Lucy; TAVARES, Felipe. “Multi-dimensional urban segregation in João
Pessoa, a coastal Brazilian northeastern city.” Cities, v. 151, 2024, p. 105-161.
FANON, Frantz. The Wretched of the Earth. New York: Grove Press, 1961.
FAUSTO, Boris. A Concise History of Brazil. Cambridge: Cambridge University
Press, 2000.
HARVEY, David. Social Justice and the City. Athens: University of Georgia Press, 2009.
FLORES, Carolina. “Advances in Research Methods for the Study of Urban
Segregation.” In: ROBERTS, Bryan R.; WILSON, Robert H. (Eds.). Urban Segregation and
Governance in the Americas. New York: Palgrave Macmillan, 2009, p. 21-35.
FLORES, Carolina; WILSON, Robert H. “Changing Patterns of Residential Segregation in
Austin.” In: ROBERTS, Bryan R.; WILSON, Robert H. (Eds.). Urban Segregation and Governance
in the Americas. New York: Palgrave Macmillan, 2009, p. 187-203.
FREYRE, Gilberto. The Masters and the Slaves: A Study in the Development of
Brazilian Civilization. New York: Knopf, 1987.
FUNARI, Armando Palermo. “Padrão de Concentração de Estruturas em São
Paulo – Quais Perspectivas para a Transformação Urbana?” In: FIX, Mariana; PEREIRA,
https://www.periodicos.unimontes.br/index.php/rds/
153
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
Paulo Romeiro; TANAKA, Gisela (Orgs.). Eixos de Estruturação da Transformação
Urbana: inovação e avaliação em São Paulo. São Paulo: Letra Capital, 2019, p. 53-100.
GIBBONS, Joseph; YANG, Tse-Chuan; BRAULT, Elizabeth; BARTON, Michael.
“Evaluating residential segregation’s relation to the clustering of poor health across
American cities.” International Journal of Environmental Research and Public Health, v.
17, n. 3910, 2020.
GROISMAN, Fernando; SUÁREZ, Ana Lourdes. “Residential Segregation in Greater
Buenos Aires.” In: ROBERTS, Bryan R.; WILSON, Robert H. (Eds.). Urban Segregation and
Governance in the Americas. New York: Palgrave Macmillan, 2009, p. 39-54.
HADJ KILANI, Bochra; DHAHER, Najem. “K-Means Clustering Algorithms in Urban
Studies: A Review of Unsupervised Machine Learning Techniques.” Authorea, 21 out. 2024.
DOI: 10.22541/au.170537264.49598441/v2.
HOLLOWAY, Thomas H. Immigrants on the Land: Coffee and Society in São
Paulo, 1886-1934. Chapel Hill: University of North Carolina Press, 1980.
IBGE (Instituto Brasileiro de Geografia e Estatística). São Paulo: panorama.
Disponível em: https://cidades.ibge.gov.br/brasil/sp/panorama. Acesso em:
02.11.2024
ISARD, Walter. Location and space-economy: a general theory relating to
industrial location, market areas, land use, trade, and urban structure. 1956.
INSTITUTO SOU DA PAZ. Mortes cometidas por policiais em serviço
aumentaram 78,5% nos primeiros oito meses de 2024; revela levantamento do Sou da
Paz. Instituto Sou da Paz. Retrieved on January 21, 2025, from
https://soudapaz.org/noticias/mortes-cometidas-por-policiais-em-servico-
aumentaram-785-nos-primeiros-oito-meses-de-2024-revela-levantamento-do-
instituto-sou-da-paz, 2024.
LANNA, Ana Lúcia Duarte. A Cidade e o Campo: O Processo de Urbanização de
São Paulo (1880-1920). São Paulo: Hucitec, 1996.
LEFEBVRE, Henri. The Production of Space. Oxford: Blackwell, 1991.
LEVINE, Robert M. The History of Brazil. Westport: Greenwood Press, 1999.
LIPSITZ, George. The Possessive Investment in Whiteness: How White People Profit
from Identity Politics. Philadelphia: Temple University Press, 2006.
LEFEBVRE, Henri. The Production of Space. Oxford: Blackwell, 1991.
LI, Zhigang; GOU, Feicui. “Residential segregation of rural migrants in post-reform
urban China.” In: MUSTERD, Sako (Ed.). Handbook of Urban Segregation. Cheltenham: Edward
Elgar, 2020, p. 55-75.
MACQUEEN, J. B. “Some Methods for Classification and Analysis of Multivariate
Observations.” In: Proceedings of the Fifth Berkeley Symposium on Mathematical
Statistics and Probability, v. 1, p. 281-297, 1967.
MIGNOLO, Walter D. The Darker Side of Western Modernity: Global Futures,
Decolonial Options. Durham: Duke University Press, 2011.
MARQUES, Eduardo; FRANÇA, Danilo. “Segregation by class and race in São Paulo.” In:
MUSTERD, Sako (Ed.). Handbook of Urban Segregation. Cheltenham: Edward Elgar, 2020, p.
36-54.
MARQUES, Eduardo; RODRIGUES, Leandro. “O Programa Minha Casa Minha Vida na
metrópole paulistana: atendimento habitacional e padrões de segregação.” Revista Brasileira
de Estudos Urbanos e Regionais, v. 15, n. 2, p. 159-177, novembro 2013.
MARTINS, José de Souza. Fronteira: A Degradação do Outro nos Confins do
Humano. São Paulo: Hucitec, 2000.
https://www.periodicos.unimontes.br/index.php/rds/
154
Revista Desenvolvimento Social, vol. 30, n. 2, jul/dez, 2024
PPGDS/Unimontes-MG
MORAN, Patrick A. P. “Notes on Continuous Stochastic Phenomena.”
Biometrika, v. 37, n. 1-2, p. 17-23, 1950.
NIELSEN, Michael M.; HENNERDAL, Pontus. “Changes in the residential
segregation of immigrants in Sweden from 1990 to 2012: Using a multi-scalar
segregation measure that accounts for the modifiable areal unit problem.” Applied
Geography, v. 87, p. 73–84, 2017.
OBSERVA SAUDE (Observatório de Saúde da Região Metropolitana de São
Paulo). Internações Hospitalares. Disponível em:
https://www.fsp.usp.br/observasaude/internacoes-hospitalares. Acesso em: 3 nov.
2024.
POULSEN, Michael; JOHNSTON, Ron; FORREST, James. “The intensity of ethnic
residential clustering: Exploring scale effects using local indicators of spatial
association.” Environment & Planning A, v. 42, n. 4, p. 874–894, 2010.
QUIJANO, Aníbal. Coloniality of Power, Eurocentrism, and Latin America. Nepantla:
Views from South, v. 1, n. 3, p. 533–580, 2000.
PRÉTECEILLE, Edmond; CARDOSO, Adalberto. “Socioeconomic segregation and the
middle classes in Paris, Rio de Janeiro and São Paulo: a comparative perspective.” In:
MUSTERD, Sako (Ed.). Handbook of Urban Segregation. Cheltenham: Edward Elgar, 2020, p.
270-288.
RIBEIRO, Arthur Duarte. Estado de São Paulo: carta organizada com os mais
recentes documentos. 1924. Biblioteca Digital Luso-Brasileira. Disponível em:
https://bdlb.bn.gov.br/acervo/handle/ 20.500.12156.3/44500. Acesso em: 03.11.2024
ROUSSEEUW, Peter J. “Silhouettes: a graphical aid to the interpretation and
validation of cluster analysis.” Journal of Computational and Applied Mathematics, v.
20, p. 53-65, 1987. North-Holland.
SANTOS, Boaventura de Sousa. Epistemologies of the South: Justice Against
Epistemicide. London: Routledge, 2014.
SASSEN, Saskia. The Global City: New York, London, Tokyo. Princeton: Princeton
University Press, 2001.
SOJA, Edward W. Postmetropolis: Critical Studies of Cities and Regions. Oxford:
Blackwell, 2000.
TAJFEL, Henri; TURNER, John. “The Social Identity Theory of Intergroup
Behavior.” In: WORCHEL, Stephen; AUSTIN, William G. (Eds.). Psychology of Intergroup
Relations. Chicago: Nelson-Hall, 1979. p. 33-47.
THORNDIKE, R. L. “Who belongs in the family?.” Psychometrika, v. 18, p. 267-
276, 1953.
TORRES, Haroldo da Gama; BICHIR, Renata Mirandola. “Residential Segregation
in São Paulo: Consequences for Urban Policies.” In: ROBERTS, Bryan R.; WILSON,
Robert H. (Eds.). Urban Segregation and Governance in the Americas. New York:
Palgrave Macmillan, 2009, p. 145-165.
TURNER, John C. Rediscovering the Social Group: A Self-Categorization Theory.
Oxford: Blackwell, 1987.
VILLAÇA, Flávio. “São Paulo: Urban segregation and inequality.” Estudos
Avançados, v. 25, n. 71, p. 37-58, 2011.
WACQUANT, Loic. Urban Outcasts: A Comparative Sociology of Advanced Marginality.
Cambridge: Polity Press, 2008.
WEISBERG, Sanford. Applied Linear Regression. Hoboken: John Wiley & Sons,
2005.
